diff --git a/.DS_Store b/.DS_Store index 1a63846e..91e4c33d 100644 Binary files a/.DS_Store and b/.DS_Store differ diff --git a/.claude/commands/wiki-graph.md b/.claude/commands/wiki-graph.md new file mode 100644 index 00000000..64c0338d --- /dev/null +++ b/.claude/commands/wiki-graph.md @@ -0,0 +1,18 @@ +Build the LLM Wiki knowledge graph. + +Usage: /wiki-graph + +First try running: python tools/build_graph.py --open + +If that fails (missing dependencies), build the graph manually: + +1. Use Grep to find all [[wikilinks]] across every file in wiki/ +2. Build a nodes list: one node per wiki page, with id=relative-path, label=title, type from frontmatter +3. Build an edges list: one edge per [[wikilink]], tagged EXTRACTED +4. Infer additional implicit relationships between pages not captured by wikilinks — tag these INFERRED with a confidence score (0.0–1.0); tag low-confidence ones AMBIGUOUS +5. Write graph/graph.json with {nodes, edges, built: today} +6. Write graph/graph.html as a self-contained vis.js page (nodes colored by type, edges colored by type, interactive, searchable) + +After building, summarize: node count, edge count, breakdown by type, and the most connected nodes (hubs). + +Append to wiki/log.md: ## [today's date] graph | Knowledge graph rebuilt diff --git a/.claude/commands/wiki-ingest.md b/.claude/commands/wiki-ingest.md new file mode 100644 index 00000000..f8e3a37c --- /dev/null +++ b/.claude/commands/wiki-ingest.md @@ -0,0 +1,18 @@ +Ingest a source document into the LLM Wiki. + +Usage: /wiki-ingest $ARGUMENTS + +$ARGUMENTS should be the path to a file in raw/, e.g. `raw/articles/my-article.md` + +Follow the Ingest Workflow defined in CLAUDE.md exactly: +1. Read the source file at the given path +2. Read wiki/index.md and wiki/overview.md for current context +3. Write wiki/sources/.md (source page format per CLAUDE.md) +4. Update wiki/index.md — add the new entry under Sources +5. Update wiki/overview.md — revise synthesis if warranted +6. Create/update entity pages (wiki/entities/) for key people, companies, projects +7. Create/update concept pages (wiki/concepts/) for key ideas and frameworks +8. Flag any contradictions with existing wiki content +9. Append to wiki/log.md: ## [today's date] ingest | + +After completing all writes, summarize: what was added, which pages were created or updated, and any contradictions found. diff --git a/.claude/commands/wiki-lint.md b/.claude/commands/wiki-lint.md new file mode 100644 index 00000000..fe45b691 --- /dev/null +++ b/.claude/commands/wiki-lint.md @@ -0,0 +1,19 @@ +Health-check the LLM Wiki for issues. + +Usage: /wiki-lint + +Follow the Lint Workflow defined in CLAUDE.md: + +Structural checks (use Grep and Glob tools): +1. Orphan pages — wiki pages with no inbound [[wikilinks]] from other pages +2. Broken links — [[WikiLinks]] pointing to pages that don't exist +3. Missing entity pages — names referenced in 3+ pages but lacking their own page + +Semantic checks (read and reason over page content): +4. Contradictions — claims that conflict between pages +5. Stale summaries — pages not updated after newer sources changed the picture +6. Data gaps — important questions the wiki can't answer; suggest specific sources to find + +Output a structured markdown lint report. At the end, ask if the user wants it saved to wiki/lint-report.md. + +Append to wiki/log.md: ## [today's date] lint | Wiki health check diff --git a/.claude/commands/wiki-query.md b/.claude/commands/wiki-query.md new file mode 100644 index 00000000..8a6a6df1 --- /dev/null +++ b/.claude/commands/wiki-query.md @@ -0,0 +1,14 @@ +Query the LLM Wiki and synthesize an answer. + +Usage: /wiki-query $ARGUMENTS + +$ARGUMENTS is the question to answer, e.g. `What are the main themes across all sources?` + +Follow the Query Workflow defined in CLAUDE.md: +1. Read wiki/index.md to identify the most relevant pages +2. Read those pages (up to ~10 most relevant) +3. Synthesize a thorough markdown answer with [[PageName]] wikilink citations +4. Include a ## Sources section at the end listing pages you drew from +5. Ask the user if they want the answer saved as wiki/syntheses/<slug>.md + +If the wiki is empty, say so and suggest running /wiki-ingest first. diff --git a/CLAUDE(ENGLISH).md b/CLAUDE(ENGLISH).md new file mode 100644 index 00000000..345219f7 --- /dev/null +++ b/CLAUDE(ENGLISH).md @@ -0,0 +1,230 @@ +# LLM Wiki Agent — Schema & Workflow Instructions + +This wiki is maintained entirely by Claude Code. No API key or Python scripts needed — just open this repo in Claude Code and talk to it. + +## Slash Commands (Claude Code) + +| Command | What to say | +|---|---| +| `/wiki-ingest` | `ingest raw/my-article.md` | +| `/wiki-query` | `query: what are the main themes?` | +| `/wiki-lint` | `lint the wiki` | +| `/wiki-graph` | `build the knowledge graph` | + +Or just describe what you want in plain English: +- *"Ingest this file: raw/papers/attention-is-all-you-need.md"* +- *"What does the wiki say about transformer models?"* +- *"Check the wiki for orphan pages and contradictions"* +- *"Build the graph and show me what's connected to RAG"* + +Claude Code reads this file automatically and follows the workflows below. + +--- + +## Directory Layout + +``` +raw/ # Immutable source documents — never modify these +wiki/ # Claude owns this layer entirely + index.md # Catalog of all pages — update on every ingest + log.md # Append-only chronological record + overview.md # Living synthesis across all sources + sources/ # One summary page per source document + entities/ # People, companies, projects, products + concepts/ # Ideas, frameworks, methods, theories + syntheses/ # Saved query answers +graph/ # Auto-generated graph data +tools/ # Optional standalone Python scripts (require ANTHROPIC_API_KEY) +``` + +--- + +## Page Format + +Every wiki page uses this frontmatter: + +```yaml +--- +title: "Page Title" +type: source | entity | concept | synthesis +tags: [] +sources: [] # list of source slugs that inform this page +last_updated: YYYY-MM-DD +--- +``` + +Use `[[PageName]]` wikilinks to link to other wiki pages. + +--- + +## Ingest Workflow + +Triggered by: *"ingest <file>"* or `/wiki-ingest` + +Steps (in order): +1. Read the source document fully using the Read tool +2. Read `wiki/index.md` and `wiki/overview.md` for current wiki context +3. Write `wiki/sources/<slug>.md` — use the source page format below +4. Update `wiki/index.md` — add entry under Sources section +5. Update `wiki/overview.md` — revise synthesis if warranted +6. Update/create entity pages for key people, companies, projects mentioned +7. Update/create concept pages for key ideas and frameworks discussed +8. Flag any contradictions with existing wiki content +9. Append to `wiki/log.md`: `## [YYYY-MM-DD] ingest | <Title>` + +### Source Page Format + +```markdown +--- +title: "Source Title" +type: source +tags: [] +date: YYYY-MM-DD +source_file: raw/... +--- + +## Summary +2–4 sentence summary. + +## Key Claims +- Claim 1 +- Claim 2 + +## Key Quotes +> "Quote here" — context + +## Connections +- [[EntityName]] — how they relate +- [[ConceptName]] — how it connects + +## Contradictions +- Contradicts [[OtherPage]] on: ... +``` + +### Domain-Specific Templates + +If the source falls into a specific domain (e.g., personal diary, meeting notes), the agent should use a specialized template instead of the default generic one above: + +#### Diary / Journal Template +```markdown +--- +title: "YYYY-MM-DD Diary" +type: source +tags: [diary] +date: YYYY-MM-DD +--- +## Event Summary +... +## Key Decisions +... +## Energy & Mood +... +## Connections +... +## Shifts & Contradictions +... +``` + +#### Meeting Notes Template +```markdown +--- +title: "Meeting Title" +type: source +tags: [meeting] +date: YYYY-MM-DD +--- +## Goal +... +## Key Discussions +... +## Decisions Made +... +## Action Items +... +``` + +--- + +## Query Workflow + +Triggered by: *"query: <question>"* or `/wiki-query` + +Steps: +1. Read `wiki/index.md` to identify relevant pages +2. Read those pages with the Read tool +3. Synthesize an answer with inline citations as `[[PageName]]` wikilinks +4. Ask the user if they want the answer filed as `wiki/syntheses/<slug>.md` + +--- + +## Lint Workflow + +Triggered by: *"lint the wiki"* or `/wiki-lint` + +Use Grep and Read tools to check for: +- **Orphan pages** — wiki pages with no inbound `[[links]]` from other pages +- **Broken links** — `[[WikiLinks]]` pointing to pages that don't exist +- **Contradictions** — claims that conflict across pages +- **Stale summaries** — pages not updated after newer sources +- **Missing entity pages** — entities mentioned in 3+ pages but lacking their own page +- **Data gaps** — questions the wiki can't answer; suggest new sources + +Output a lint report and ask if the user wants it saved to `wiki/lint-report.md`. + +--- + +## Graph Workflow + +Triggered by: *"build the knowledge graph"* or `/wiki-graph` + +When the user asks to build the graph, run `tools/build_graph.py` which: +- Pass 1: Parses all `[[wikilinks]]` → deterministic `EXTRACTED` edges +- Pass 2: Infers implicit relationships → `INFERRED` edges with confidence scores +- Runs Louvain community detection +- Outputs `graph/graph.json` + `graph/graph.html` + +If the user doesn't have Python/dependencies set up, instead generate the graph data manually: +1. Use Grep to find all `[[wikilinks]]` across wiki pages +2. Build a node/edge list +3. Write `graph/graph.json` directly +4. Write `graph/graph.html` using the vis.js template + +--- + +## Naming Conventions + +- Source slugs: `kebab-case` matching source filename +- Entity pages: `TitleCase.md` (e.g. `OpenAI.md`, `SamAltman.md`) +- Concept pages: `TitleCase.md` (e.g. `ReinforcementLearning.md`, `RAG.md`) +- Source pages: `kebab-case.md` + +## Index Format + +```markdown +# Wiki Index + +## Overview +- [Overview](overview.md) — living synthesis + +## Sources +- [Source Title](sources/slug.md) — one-line summary + +## Entities +- [Entity Name](entities/EntityName.md) — one-line description + +## Concepts +- [Concept Name](concepts/ConceptName.md) — one-line description + +## Syntheses +- [Analysis Title](syntheses/slug.md) — what question it answers +``` + +## Log Format + +Each entry starts with `## [YYYY-MM-DD] <operation> | <title>` so it's grep-parseable: + +``` +grep "^## \[" wiki/log.md | tail -10 +``` + +Operations: `ingest`, `query`, `lint`, `graph` diff --git a/CLAUDE.md b/CLAUDE.md index 345219f7..9fefa0fd 100644 --- a/CLAUDE.md +++ b/CLAUDE.md @@ -1,78 +1,133 @@ -# LLM Wiki Agent — Schema & Workflow Instructions +# LLM Wiki Agent — Schema & Workflow Instructions(中文版增强规范) -This wiki is maintained entirely by Claude Code. No API key or Python scripts needed — just open this repo in Claude Code and talk to it. +本 Wiki 完全由 Claude Code 自动维护。无需 API Key 或 Python 脚本 —— 只需在 Claude Code 中打开本仓库并与其对话。 -## Slash Commands (Claude Code) +--- +# 🔴 全局强制规则(CRITICAL) -| Command | What to say | +## 1. 输出语言(必须遵守) + +- 所有输出必须使用**简体中文** +- 专有名词允许保留英文,但首次出现必须附带中文解释 +- 禁止中英混合句(术语除外) +- 不允许输出纯英文总结或分析 + +示例: + +Transformer(变压器模型,一种基于注意力机制的神经网络架构) + +--- + +## 2. 输出风格(严格限制) + +所有输出必须: + +- 去修辞(禁止 narrative 风格) +- 去模糊(禁止“可能”“大概”等词) +- 信息密度最大化 +- 面向“知识结构化”,而非阅读体验 + +优先级: + +结构 > 关系 > 结论 > 描述 + +--- + +## 3. 结构化语义(必须) + +所有页面必须遵循结构化语义规则: + +- Summary 必须使用固定字段 +- Claim 必须符合标准语法 +- Connections 必须使用关系类型 +- 禁止自由发挥 + +--- + +# Slash Commands(Claude Code) + +| Command | 使用方式 | |---|---| -| `/wiki-ingest` | `ingest raw/my-article.md` | -| `/wiki-query` | `query: what are the main themes?` | +| `/wiki-ingest` | `ingest raw/your-file.md` | +| `/wiki-query` | `query: 你的问题` | | `/wiki-lint` | `lint the wiki` | | `/wiki-graph` | `build the knowledge graph` | -Or just describe what you want in plain English: -- *"Ingest this file: raw/papers/attention-is-all-you-need.md"* -- *"What does the wiki say about transformer models?"* -- *"Check the wiki for orphan pages and contradictions"* -- *"Build the graph and show me what's connected to RAG"* +--- + +## 自然语言示例 + +- ingest raw/papers/attention-is-all-you-need.md +- query: Transformer 的核心机制是什么? +- lint the wiki +- build the graph and analyze RAG + +Claude Code 会自动读取本文件并执行以下工作流。 + -Claude Code reads this file automatically and follows the workflows below. --- -## Directory Layout +# Directory Layout(目录结构) ``` -raw/ # Immutable source documents — never modify these -wiki/ # Claude owns this layer entirely - index.md # Catalog of all pages — update on every ingest - log.md # Append-only chronological record - overview.md # Living synthesis across all sources - sources/ # One summary page per source document - entities/ # People, companies, projects, products - concepts/ # Ideas, frameworks, methods, theories - syntheses/ # Saved query answers -graph/ # Auto-generated graph data -tools/ # Optional standalone Python scripts (require ANTHROPIC_API_KEY) -``` +raw/ # 原始文档(不可修改) +wiki/ # 知识层(由 Claude 完全维护) + index.md # 页面索引(每次 ingest 必须更新) + log.md # 追加式日志 + overview.md # 全局知识总结 + sources/ # 每个原始文档对应一个页面 + entities/ # 实体(人/公司/产品/项目) + concepts/ # 概念(方法/理论/框架) + syntheses/ # 查询结果沉淀 +graph/ # 自动生成的图数据 +tools/ # 可选 Python 工具 (require ANTHROPIC_API_KEY) +```` + --- -## Page Format +# Page Format(页面格式) -Every wiki page uses this frontmatter: +每个页面必须包含: ```yaml --- +id: unique_id title: "Page Title" type: source | entity | concept | synthesis tags: [] -sources: [] # list of source slugs that inform this page +sources: [] # 来源 last_updated: YYYY-MM-DD --- -``` +```` -Use `[[PageName]]` wikilinks to link to other wiki pages. +必须使用 `[[PageName]]` 进行链接。 --- -## Ingest Workflow +# Ingest Workflow(摄取流程) -Triggered by: *"ingest <file>"* or `/wiki-ingest` +触发方式: +- `/wiki-ingest` +- 或:`ingest <file>` -Steps (in order): -1. Read the source document fully using the Read tool -2. Read `wiki/index.md` and `wiki/overview.md` for current wiki context -3. Write `wiki/sources/<slug>.md` — use the source page format below -4. Update `wiki/index.md` — add entry under Sources section -5. Update `wiki/overview.md` — revise synthesis if warranted -6. Update/create entity pages for key people, companies, projects mentioned -7. Update/create concept pages for key ideas and frameworks discussed -8. Flag any contradictions with existing wiki content -9. Append to `wiki/log.md`: `## [YYYY-MM-DD] ingest | <Title>` +--- -### Source Page Format +## 执行步骤(严格顺序) +1. 使用 Read 工具完整读取 source 文档 +2. 读取 `wiki/index.md` 和 `wiki/overview.md` +3. 生成 `wiki/sources/<slug>.md` +4. 更新 `wiki/index.md` +5. 更新 `wiki/overview.md`(如有必要) +6. 创建或更新 Entity 页面 +7. 创建或更新 Concept 页面 +8. 检测并记录冲突 +9. 追加 `wiki/log.md` + +--- + +# Source Page Format(增强结构) ```markdown --- @@ -84,28 +139,40 @@ source_file: raw/... --- ## Summary -2–4 sentence summary. +- 核心主题: +- 问题域: +- 方法/机制: +- 结论/价值: ## Key Claims -- Claim 1 -- Claim 2 +- (必须符合:主体 + 机制 + 结果) ## Key Quotes -> "Quote here" — context +> "引用内容" — 上下文说明 + +## Key Concepts +- [[ConceptName]]:定义 + +## Key Entities +- [[EntityName]]:角色说明 ## Connections -- [[EntityName]] — how they relate -- [[ConceptName]] — how it connects +- [[A]] ← depends_on ← [[B]] +- [[C]] ← extends ← [[D]] ## Contradictions -- Contradicts [[OtherPage]] on: ... +- 与 [[OtherPage]] 冲突: + - 冲突点: + - 当前观点: + - 对方观点: ``` -### Domain-Specific Templates +--- -If the source falls into a specific domain (e.g., personal diary, meeting notes), the agent should use a specialized template instead of the default generic one above: +# Domain-Specific Templates(领域模板) + +## Diary / Journal -#### Diary / Journal Template ```markdown --- title: "YYYY-MM-DD Diary" @@ -114,18 +181,16 @@ tags: [diary] date: YYYY-MM-DD --- ## Event Summary -... ## Key Decisions -... ## Energy & Mood -... ## Connections -... ## Shifts & Contradictions -... ``` -#### Meeting Notes Template +--- + +## Meeting Notes + ```markdown --- title: "Meeting Title" @@ -134,97 +199,152 @@ tags: [meeting] date: YYYY-MM-DD --- ## Goal -... ## Key Discussions -... ## Decisions Made -... ## Action Items -... ``` --- -## Query Workflow +# Entity & Concept Rules(关键增强) -Triggered by: *"query: <question>"* or `/wiki-query` +## Entity(实体) -Steps: -1. Read `wiki/index.md` to identify relevant pages -2. Read those pages with the Read tool -3. Synthesize an answer with inline citations as `[[PageName]]` wikilinks -4. Ask the user if they want the answer filed as `wiki/syntheses/<slug>.md` +创建条件: +- 出现 ≥ 2 次 + 或 +- 对主题有关键影响 + +类型: +- 人 / 公司 / 产品 / 项目 --- -## Lint Workflow +## Concept(概念) +创建条件: +- 可抽象 +- 可复用 +- 非具体实例 +--- -Triggered by: *"lint the wiki"* or `/wiki-lint` +## 命名规范(强制) +- 使用唯一标准名称 +- 所有别名写入页面: -Use Grep and Read tools to check for: -- **Orphan pages** — wiki pages with no inbound `[[links]]` from other pages -- **Broken links** — `[[WikiLinks]]` pointing to pages that don't exist -- **Contradictions** — claims that conflict across pages -- **Stale summaries** — pages not updated after newer sources -- **Missing entity pages** — entities mentioned in 3+ pages but lacking their own page -- **Data gaps** — questions the wiki can't answer; suggest new sources - -Output a lint report and ask if the user wants it saved to `wiki/lint-report.md`. +```markdown +## Aliases +- GPT4 +- GPT-4 +``` --- -## Graph Workflow +## 去重机制(必须) -Triggered by: *"build the knowledge graph"* or `/wiki-graph` - -When the user asks to build the graph, run `tools/build_graph.py` which: -- Pass 1: Parses all `[[wikilinks]]` → deterministic `EXTRACTED` edges -- Pass 2: Infers implicit relationships → `INFERRED` edges with confidence scores -- Runs Louvain community detection -- Outputs `graph/graph.json` + `graph/graph.html` - -If the user doesn't have Python/dependencies set up, instead generate the graph data manually: -1. Use Grep to find all `[[wikilinks]]` across wiki pages -2. Build a node/edge list -3. Write `graph/graph.json` directly -4. Write `graph/graph.html` using the vis.js template +创建前必须: +1. 搜索 index +2. 判断是否存在 +3. 存在则更新 --- -## Naming Conventions +# Query Workflow(查询流程) -- Source slugs: `kebab-case` matching source filename -- Entity pages: `TitleCase.md` (e.g. `OpenAI.md`, `SamAltman.md`) -- Concept pages: `TitleCase.md` (e.g. `ReinforcementLearning.md`, `RAG.md`) -- Source pages: `kebab-case.md` +触发: +- `/wiki-query` +- 或:`query: 问题` -## Index Format +--- + +## 步骤 + +1. 读取 index +2. 找到相关页面 +3. 使用 Read 工具加载 +4. 输出结构化答案 +5. 使用 `[[Page]]` 引用 +6. 询问是否保存为 synthesis + +--- + +# Lint Workflow(校验) + +检查内容: + +- 孤立页面 +- 断链 +- 冲突 +- 过期内容 +- 缺失实体 +- 知识空白 + +--- + +# Graph Workflow(知识图谱) + +触发: +- `/wiki-graph` + +--- + +执行: +- 优先运行 `tools/build_graph.py` +- 否则手动构建: + +步骤: +1. 提取所有 `[[links]]` +2. 构建节点与边 +3. 输出 `graph.json` + +--- + +# Naming Conventions(命名规范) +- Source:kebab-case +- Entity:TitleCase +- Concept:TitleCase + +--- + +# Index Format(索引结构) ```markdown # Wiki Index ## Overview -- [Overview](overview.md) — living synthesis +- [Overview](overview.md) ## Sources -- [Source Title](sources/slug.md) — one-line summary +- [Title](sources/slug.md) ## Entities -- [Entity Name](entities/EntityName.md) — one-line description +- [Entity](entities/Entity.md) ## Concepts -- [Concept Name](concepts/ConceptName.md) — one-line description +- [Concept](concepts/Concept.md) ## Syntheses -- [Analysis Title](syntheses/slug.md) — what question it answers +- [Title](syntheses/slug.md) ``` -## Log Format +--- -Each entry starts with `## [YYYY-MM-DD] <operation> | <title>` so it's grep-parseable: +# Log Format(日志) ``` -grep "^## \[" wiki/log.md | tail -10 +## [YYYY-MM-DD] ingest | 标题 ``` -Operations: `ingest`, `query`, `lint`, `graph` +--- + +# ✅ 最终目标 + +该系统用于: + +- 知识沉淀 +- 结构化理解 +- 自动图谱构建 +- Agent 推理支持 + +--- + +# END \ No newline at end of file diff --git a/openclaw/knowledgebase/公众号爆款文章Skill-冰河.md b/openclaw/knowledgebase/公众号爆款文章Skill-冰河.md deleted file mode 100644 index 631ec1d3..00000000 --- a/openclaw/knowledgebase/公众号爆款文章Skill-冰河.md +++ /dev/null @@ -1,227 +0,0 @@ ---- -title: 公众号爆款文章制造机:这套 Skill 工具,是我送给你最硬核的赚钱外挂! -source: -author: shenwei -published: -created: -description: -tags: [] ---- - -# 公众号爆款文章制造机:这套 Skill 工具,是我送给你最硬核的赚钱外挂! - -> 来源: X (Twitter) @冰河 -> 发布时间: 2026年3月12日 23:30 -> 原文链接: https://x.com/binghe/status/2032116993378435433 - ---- - -容我先吹吹牛逼!公众号内容创业这条路,我们走了五年了,矩阵运营了多个帐号,有了些小小的成绩,下面的内容全是多年精华经验浓缩! - -文章长,内容干!如果你看不下去,那就帮我点点赞吧!感谢! - -> 注:本篇文章 Skill 已经发布在"钱来有道"社群,欢迎加入,全年免费更新,众多赚钱类教程任意阅读!性感的我会一对一服务!机会难得,广告结束! 哦对了!本 Skill 也会发布在订阅区!欢迎订阅! - ---- - -公众号赛道,一直都是我认为最低成本,最容易得到正反馈的选择。 - -只要你会写作,或者"不那么"会写,都能取得成绩。 - -还记得以前我写过一篇文章:手把手教你如何用AI写爆款文章,轻松日赚到2000块! - -很多人通过这篇文章认识了我,加入了社群! - -一年过去了,公众号赛道的文章类型和平台政策已经大变样了,我决定在一年后重新把以前的经验整理成 Skill,给大家一个新的选择。 - -好,确定了写公众号这个低成本创收的赛道,那你平时在写作的时候会经常遇到以下几个问题: - -(配图) - -这些问题,几乎每个做公众号的人都经历过。 - -今天这个 Skill 我需要它扮演的角色不是个写稿机器 ,而应该是一个会思考的编辑! - -市面上的 AI 写作工具,大多数只能做一件事:你说写什么,它就写什么。 - -但写公众号爆款,从来不是「有文字输出」就够了。 - -你需要有人帮你判断选题够不够刁钻,需要有人帮你想清楚这件事从哪个角度切进去文字最有力量,需要有人在你想法还模糊的时候,把它描绘清晰——而不是催你「请提供更多信息」。 - -这就是这个 Skill 要做的事! - ---- - -## 实操来讲解这个 Skill 它帮你解决了哪些真实的痛点? - -### 第一步 - -先把 Skill 文件放到全局目录中的"skills "里,然后在其它地方新建目录,在当前目录下打开 Claude Code - -让它调用 Skill ,开始写作流程: - -(配图) - -可以看到,skill 没有开始就准备写,而是帮你写之前整理写作逻辑,结构,和切入角度入手! - -比如我直接说:我想写关于娱乐八卦 - -它调用的就是娱乐类文章写法的 skill设定内容! - -但其实这个 Skill有好多写作种类的设定: - -(配图) - -不管你选择哪个场景,都会有相对应的 Skill 进行规则设定,让你写的内容更符合当前分类爆款内容结构。 - -我们下面是拿娱乐八卦来做一下演示。 - -(配图) - -请仔细看看这种分类下的文章结构,这是我拆解了近百篇娱乐类公众号文章提炼的写作精华内容。 - -下面你可以根据提示,完成下一步具体的写作要求。 - -比如某个娱乐事件,这就是要求输入主题内容了。 - -> 一般情况,我强烈建议你多上传一些关于这个主题内容的其它不同类文章,语料越多,写出来的内容越丰富,AI 也会综合判断什么样的内容结构排列组合会更有针对性。这样就从根上避免了抄袭!因为这已经是一篇综合多方信息源的原创文章了。 - -案例操作,简单点!我让他查了查薛之谦最近的娱乐新闻,准备写关于他的内容。 - -(配图) - -同一个热点,一百个人在写! - -大多数人写的角度都差不多,读者看了开头就知道后面是什么。 - -这个 Skill 有一个专门的 `~选题` 功能。 - -你给它一个方向,它帮你找那个「别人没想到,但你一说出来就觉得绝」的切入角度。 - -每个角度它都会告诉你爆点在哪里、为什么这个方向有传播潜力,让你在动笔前就心里有数。 - -(配图) - -可以看到,它其实给了我很多个选题(一共五个),每个选题的文章方向、爆点在哪里全部进行了规划! - -你从这些规划中参考一下,是否有一些是别人没讲过的,或者是几个选题排列组合一下成为一个新的内容结构,都是可以的,要学会活学活用! - -(配图) - -我随便先了"角度五",到这里这个 Skill 也没有开始写!这很关键! - -我们要先看大纲,每部分的内容规划如何?最后的总字数是否满足你的要求? - -核心观点是否寓意突出? - -在做了综合判断后,再开始写正文! - -(配图) - -注意:这里如果你不满意都可随时调整,我把文章字数设定了与写作分类相关,并不是越多越好!但字数少,也不能表现出文章深度! - -输入写正文后,产出内容,并同时保存在当前项目目录里! - -这里我要说明几个事情: - -1. 如果你使用的模型是国产的话,文章内容非常 AI 化,格式化(喜欢用:第一,第二,不是而是!)一定要手动润色!Claude 的opus 模型就没有这个问题。所以要注意! - -2. 内容结构完全是依照娱乐类文章的写作模式在编写。调动情绪,引导评论,制造互动性,如下图结尾部分: - -(配图) - -3. 所有内容为.md 格式,你可以上网找一些转成公众号的编辑器,直接粘贴进去,就成了公众号文章内容了。我们以前有过分享! - -4. 要审稿,别直接发布!每次产出的内容必审稿,AI 不是万能的,你要去对内容中的情绪、事件完整性、数据准确度进行审核,别闹出笑话! - -(配图) - -以前是先有标题,后有内容,但这次我把这个顺序改了一下,先有内容后产标题! - -这样标题的生成就会更有针对性,我们可以输入"~标题",它会调用 Skill 中的标题公式,进行有目的性的产出 10 条备选标题内容! - -以 AI 的角度提供3 条选择,当然最终选择权在你这里,如果不满意继续让它产出! - -我个人认为这些标题还是非常有"娱乐性"的,他不是标题党,而是情绪先行! - -(配图) - -选择完标题,自动加入到文章中,然后就到了精细打磨阶段了! - -其中有两个指令非常重要就是:"~改稿","~精写" - -其实就是哪里不对改哪里,你只需要复制某个段落,或者是某部分内容,然后再加上指令,就可指哪打哪儿!也可以直接改语气,反正只要是修改,使用这两个命令就完了! - -这样修改出来的内容是会根据上下文进行改写,不会偏离主题!而不需要全部重写,提升效率! - -这里就不做演示了! - -最后写完,我们也要分析一下文章到底写的怎么样?所谓当局者迷! - -那好,输入指令"~分析" - -(配图) - -(配图) - -分析报告非常详细,到时你自己看吧!如果你觉得它说的有道理,那好就让他继续修改! - -如果没有问题就可以继续产出了。 - -可以看出来,在写之前,我们一直做的是陪伴编辑的角色,这就是这个 Skill 的意义,知其然,也要知其所以然! - -不然,你只是个使用工具无情的机器! - -(配图) - -这就结束了吗? - -不,当然没有! - -这套 Skill 还有个非常牛的设定:不管你写的是什么内容类型,除了大类规则套用,还可以跨类型通用规则适配! - -(配图) - -我把常见的通用类型也在规则在做了设计,不管是常见的不常见的,只要写好通用规则,自动启动爆款格式,进行有人味,有立场,有冲突,有情绪,有互动的文章写作逻辑! - -当你想用自己的风格写作,当然也可以! - -我们在指令中还加入了一个"~风格" 指令,输入这个指令的时候,给你四种选择: - -(配图) - -A. 查看当前已经写好的文章风格,以后沿用(前提是你喜欢这个笔风) - -B. 直接告诉 AI 你就要什么风格,因为你也没想好。 - -C. 上传你的参考文章,他会提炼风格特征进行模仿,产出属于你自己味道的内容。(上传文件最好是.md 格式,多传一些也没问题。) - -D. 保持现状 - -是不是感觉这套 Skill 有点意思了? - -哈哈哈,越灵活的操作越可以发挥它的最大功效!我们不能把自己的思维固住! - -可能这么多指令你记不住 ,没关系! - -只要输入"~指令",就会把所有指令都做一展示,还有指令功能进行说明。我是不是很贴心? - -(配图) - -好了,这个 Skill 其实并不复杂,只是里面的写作规则记录的比较丰富! - -我们在面对任何写作领域的时候,不管是公众号,推特,头条其它写文章的平台都可以适用! - -但好的文章要有温度,有共情,有特别的切入角度,并且能引起讨论,这才是一篇合格的文章! - -内容的深度,关键在于你提代的语料是否足够,还与你对信息获取的能力息息相关! - ---- - -> 爱技术,爱赚钱,爱分享! -> 分享 赚钱项目 |打破信息壁垒 |副业避坑 -> 能拆解的赚钱项目才有意义 -> 知识星球社群/公众号:"钱来有道" -> 小宇宙 |Podcasts |QQ音乐 播客 "钱来有道" -> 搞钱训练营 -> qianlai.flowus.cn diff --git a/openclaw/knowledgebase/夸克资源-2026-04-04.md b/openclaw/knowledgebase/夸克资源-2026-04-04.md deleted file mode 100644 index 741587db..00000000 --- a/openclaw/knowledgebase/夸克资源-2026-04-04.md +++ /dev/null @@ -1,34 +0,0 @@ ---- -title: 夸克资源 -source: -author: shenwei -published: -created: -description: -tags: [] ---- - -# 夸克资源 - -> 来源: https://t.me/quarkF/311117 -> 添加时间: 2026-04-04 - -## 资源链接 -https://t.me/quarkF/311117 - -## 频道信息 -- **频道名**:付费资源分享 -- **频道ID**:@quarkF -- **链接**:https://t.me/quarkF/311117 - -## 频道简介 -分享一些我购买的付费课程! -防失联付费资源共享群 @zhishisou -免费加入内部群,所有付费资源免费共享! -网络上能看到的,只是冰山一角。 -想解锁更真实、更硬核的内容?来群里搜搜看,也许能发现不一样的世界。 - -@tangxingvlog 可以搜索全网任何资源 - -## 内容状态 -⚠️ 频道内容需登录 Telegram 才能查看完整详情,公开链接仅显示频道简介和封面图 diff --git a/openclaw/usecases/aionui-cowork-desktop.md b/raw/Agent/usecases/aionui-cowork-desktop.md similarity index 100% rename from openclaw/usecases/aionui-cowork-desktop.md rename to raw/Agent/usecases/aionui-cowork-desktop.md diff --git a/openclaw/usecases/arxiv-paper-reader.md b/raw/Agent/usecases/arxiv-paper-reader.md similarity index 100% rename from openclaw/usecases/arxiv-paper-reader.md rename to raw/Agent/usecases/arxiv-paper-reader.md diff --git a/openclaw/usecases/autonomous-game-dev-pipeline.md b/raw/Agent/usecases/autonomous-game-dev-pipeline.md similarity index 100% rename from openclaw/usecases/autonomous-game-dev-pipeline.md rename to raw/Agent/usecases/autonomous-game-dev-pipeline.md diff --git a/openclaw/usecases/autonomous-project-management.md b/raw/Agent/usecases/autonomous-project-management.md similarity index 100% rename from openclaw/usecases/autonomous-project-management.md rename to raw/Agent/usecases/autonomous-project-management.md diff --git a/openclaw/usecases/content-factory.md b/raw/Agent/usecases/content-factory.md similarity index 100% rename from openclaw/usecases/content-factory.md rename to raw/Agent/usecases/content-factory.md diff --git a/openclaw/usecases/custom-morning-brief.md b/raw/Agent/usecases/custom-morning-brief.md similarity index 100% rename from openclaw/usecases/custom-morning-brief.md rename to raw/Agent/usecases/custom-morning-brief.md diff --git a/openclaw/usecases/daily-reddit-digest.md b/raw/Agent/usecases/daily-reddit-digest.md similarity index 100% rename from openclaw/usecases/daily-reddit-digest.md rename to raw/Agent/usecases/daily-reddit-digest.md diff --git a/openclaw/usecases/daily-youtube-digest.md b/raw/Agent/usecases/daily-youtube-digest.md similarity index 100% rename from openclaw/usecases/daily-youtube-digest.md rename to raw/Agent/usecases/daily-youtube-digest.md diff --git a/openclaw/usecases/dynamic-dashboard.md b/raw/Agent/usecases/dynamic-dashboard.md similarity index 100% rename from openclaw/usecases/dynamic-dashboard.md rename to raw/Agent/usecases/dynamic-dashboard.md diff --git a/openclaw/usecases/earnings-tracker.md b/raw/Agent/usecases/earnings-tracker.md similarity index 100% rename from openclaw/usecases/earnings-tracker.md rename to raw/Agent/usecases/earnings-tracker.md diff --git a/openclaw/usecases/event-guest-confirmation.md b/raw/Agent/usecases/event-guest-confirmation.md similarity index 100% rename from openclaw/usecases/event-guest-confirmation.md rename to raw/Agent/usecases/event-guest-confirmation.md diff --git a/openclaw/usecases/family-calendar-household-assistant.md b/raw/Agent/usecases/family-calendar-household-assistant.md similarity index 100% rename from openclaw/usecases/family-calendar-household-assistant.md rename to raw/Agent/usecases/family-calendar-household-assistant.md diff --git a/openclaw/usecases/habit-tracker-accountability-coach.md b/raw/Agent/usecases/habit-tracker-accountability-coach.md similarity index 100% rename from openclaw/usecases/habit-tracker-accountability-coach.md rename to raw/Agent/usecases/habit-tracker-accountability-coach.md diff --git a/openclaw/usecases/health-symptom-tracker.md b/raw/Agent/usecases/health-symptom-tracker.md similarity index 100% rename from openclaw/usecases/health-symptom-tracker.md rename to raw/Agent/usecases/health-symptom-tracker.md diff --git a/openclaw/usecases/inbox-declutter.md b/raw/Agent/usecases/inbox-declutter.md similarity index 100% rename from openclaw/usecases/inbox-declutter.md rename to raw/Agent/usecases/inbox-declutter.md diff --git a/openclaw/usecases/knowledge-base-rag.md b/raw/Agent/usecases/knowledge-base-rag.md similarity index 100% rename from openclaw/usecases/knowledge-base-rag.md rename to raw/Agent/usecases/knowledge-base-rag.md diff --git a/openclaw/usecases/latex-paper-writing.md b/raw/Agent/usecases/latex-paper-writing.md similarity index 100% rename from openclaw/usecases/latex-paper-writing.md rename to raw/Agent/usecases/latex-paper-writing.md diff --git a/openclaw/usecases/local-crm-framework.md b/raw/Agent/usecases/local-crm-framework.md similarity index 100% rename from openclaw/usecases/local-crm-framework.md rename to raw/Agent/usecases/local-crm-framework.md diff --git a/openclaw/usecases/market-research-product-factory.md b/raw/Agent/usecases/market-research-product-factory.md similarity index 100% rename from openclaw/usecases/market-research-product-factory.md rename to raw/Agent/usecases/market-research-product-factory.md diff --git a/openclaw/usecases/meeting-notes-action-items.md b/raw/Agent/usecases/meeting-notes-action-items.md similarity index 100% rename from openclaw/usecases/meeting-notes-action-items.md rename to raw/Agent/usecases/meeting-notes-action-items.md diff --git a/openclaw/usecases/multi-agent-team.md b/raw/Agent/usecases/multi-agent-team.md similarity index 100% rename from openclaw/usecases/multi-agent-team.md rename to raw/Agent/usecases/multi-agent-team.md diff --git a/openclaw/usecases/multi-channel-assistant.md b/raw/Agent/usecases/multi-channel-assistant.md similarity index 100% rename from openclaw/usecases/multi-channel-assistant.md rename to raw/Agent/usecases/multi-channel-assistant.md diff --git a/openclaw/usecases/multi-channel-customer-service.md b/raw/Agent/usecases/multi-channel-customer-service.md similarity index 100% rename from openclaw/usecases/multi-channel-customer-service.md rename to raw/Agent/usecases/multi-channel-customer-service.md diff --git a/openclaw/usecases/multi-source-tech-news-digest.md b/raw/Agent/usecases/multi-source-tech-news-digest.md similarity index 100% rename from openclaw/usecases/multi-source-tech-news-digest.md rename to raw/Agent/usecases/multi-source-tech-news-digest.md diff --git a/openclaw/usecases/n8n-workflow-orchestration.md b/raw/Agent/usecases/n8n-workflow-orchestration.md similarity index 100% rename from openclaw/usecases/n8n-workflow-orchestration.md rename to raw/Agent/usecases/n8n-workflow-orchestration.md diff --git a/openclaw/usecases/overnight-mini-app-builder.md b/raw/Agent/usecases/overnight-mini-app-builder.md similarity index 100% rename from openclaw/usecases/overnight-mini-app-builder.md rename to raw/Agent/usecases/overnight-mini-app-builder.md diff --git a/openclaw/usecases/personal-crm.md b/raw/Agent/usecases/personal-crm.md similarity index 100% rename from openclaw/usecases/personal-crm.md rename to raw/Agent/usecases/personal-crm.md diff --git a/openclaw/usecases/phone-based-personal-assistant.md b/raw/Agent/usecases/phone-based-personal-assistant.md similarity index 100% rename from openclaw/usecases/phone-based-personal-assistant.md rename to raw/Agent/usecases/phone-based-personal-assistant.md diff --git a/openclaw/usecases/phone-call-notifications.md b/raw/Agent/usecases/phone-call-notifications.md similarity index 100% rename from openclaw/usecases/phone-call-notifications.md rename to raw/Agent/usecases/phone-call-notifications.md diff --git a/openclaw/usecases/podcast-production-pipeline.md b/raw/Agent/usecases/podcast-production-pipeline.md similarity index 100% rename from openclaw/usecases/podcast-production-pipeline.md rename to raw/Agent/usecases/podcast-production-pipeline.md diff --git a/openclaw/usecases/polymarket-autopilot.md b/raw/Agent/usecases/polymarket-autopilot.md similarity index 100% rename from openclaw/usecases/polymarket-autopilot.md rename to raw/Agent/usecases/polymarket-autopilot.md diff --git a/openclaw/usecases/pre-build-idea-validator.md b/raw/Agent/usecases/pre-build-idea-validator.md similarity index 100% rename from openclaw/usecases/pre-build-idea-validator.md rename to raw/Agent/usecases/pre-build-idea-validator.md diff --git a/openclaw/usecases/project-state-management.md b/raw/Agent/usecases/project-state-management.md similarity index 100% rename from openclaw/usecases/project-state-management.md rename to raw/Agent/usecases/project-state-management.md diff --git a/openclaw/usecases/second-brain.md b/raw/Agent/usecases/second-brain.md similarity index 100% rename from openclaw/usecases/second-brain.md rename to raw/Agent/usecases/second-brain.md diff --git a/openclaw/usecases/self-healing-home-server.md b/raw/Agent/usecases/self-healing-home-server.md similarity index 100% rename from openclaw/usecases/self-healing-home-server.md rename to raw/Agent/usecases/self-healing-home-server.md diff --git a/openclaw/usecases/semantic-memory-search.md b/raw/Agent/usecases/semantic-memory-search.md similarity index 100% rename from openclaw/usecases/semantic-memory-search.md rename to raw/Agent/usecases/semantic-memory-search.md diff --git a/openclaw/usecases/todoist-task-manager.md b/raw/Agent/usecases/todoist-task-manager.md similarity index 100% rename from openclaw/usecases/todoist-task-manager.md rename to raw/Agent/usecases/todoist-task-manager.md diff --git a/openclaw/usecases/x-account-analysis.md b/raw/Agent/usecases/x-account-analysis.md similarity index 100% rename from openclaw/usecases/x-account-analysis.md rename to raw/Agent/usecases/x-account-analysis.md diff --git a/openclaw/usecases/x-twitter-automation.md b/raw/Agent/usecases/x-twitter-automation.md similarity index 100% rename from openclaw/usecases/x-twitter-automation.md rename to raw/Agent/usecases/x-twitter-automation.md diff --git a/openclaw/usecases/youtube-content-pipeline.md b/raw/Agent/usecases/youtube-content-pipeline.md similarity index 100% rename from openclaw/usecases/youtube-content-pipeline.md rename to raw/Agent/usecases/youtube-content-pipeline.md diff --git a/openclaw/knowledgebase/万字保姆级教程-90天跑通一人公司模式-2026-03-29.md b/raw/Agent/万字保姆级教程-90天跑通一人公司模式-2026-03-29.md similarity index 100% rename from openclaw/knowledgebase/万字保姆级教程-90天跑通一人公司模式-2026-03-29.md rename to raw/Agent/万字保姆级教程-90天跑通一人公司模式-2026-03-29.md diff --git a/openclaw/knowledgebase/万字讲透OpenClaw-Workspace深度解析-2026-03-21.md b/raw/Agent/万字讲透OpenClaw-Workspace深度解析-2026-03-21.md similarity index 100% rename from openclaw/knowledgebase/万字讲透OpenClaw-Workspace深度解析-2026-03-21.md rename to raw/Agent/万字讲透OpenClaw-Workspace深度解析-2026-03-21.md diff --git a/raw/Technical/Cloud & DevOps/Cloud DevOp Maturity - Guideline.md b/raw/Cloud & DevOps/Cloud DevOp Maturity - Guideline.md similarity index 100% rename from raw/Technical/Cloud & DevOps/Cloud DevOp Maturity - Guideline.md rename to raw/Cloud & DevOps/Cloud DevOp Maturity - Guideline.md diff --git a/raw/Technical/Cloud & DevOps/Cloud Maturity Model A Detailed Guide For Cloud Adoption.md b/raw/Cloud & DevOps/Cloud Maturity Model A Detailed Guide For Cloud Adoption.md similarity index 100% rename from raw/Technical/Cloud & DevOps/Cloud Maturity Model A Detailed Guide For Cloud Adoption.md rename to raw/Cloud & DevOps/Cloud Maturity Model A Detailed Guide For Cloud Adoption.md diff --git a/raw/Technical/Cloud & DevOps/Cloud Operating Model Key Strategies and Best Practices.md b/raw/Cloud & DevOps/Cloud Operating Model Key Strategies and Best Practices.md similarity index 100% rename from raw/Technical/Cloud & DevOps/Cloud Operating Model Key Strategies and Best Practices.md rename to raw/Cloud & DevOps/Cloud Operating Model Key Strategies and Best Practices.md diff --git a/raw/Technical/Cloud & DevOps/DevOps Culture and Transformation Fostering Collaboration, Agile Practices, and Innovation LinkedIn.md b/raw/Cloud & DevOps/DevOps Culture and Transformation Fostering Collaboration, Agile Practices, and Innovation LinkedIn.md similarity index 100% rename from raw/Technical/Cloud & DevOps/DevOps Culture and Transformation Fostering Collaboration, Agile Practices, and Innovation LinkedIn.md rename to raw/Cloud & DevOps/DevOps Culture and Transformation Fostering Collaboration, Agile Practices, and Innovation LinkedIn.md diff --git a/raw/Technical/Cloud & DevOps/DevOps Maturity Model From Traditional IT to Advanced DevOps.md b/raw/Cloud & DevOps/DevOps Maturity Model From Traditional IT to Advanced DevOps.md similarity index 100% rename from raw/Technical/Cloud & DevOps/DevOps Maturity Model From Traditional IT to Advanced DevOps.md rename to raw/Cloud & DevOps/DevOps Maturity Model From Traditional IT to Advanced DevOps.md diff --git a/raw/Technical/Cloud & DevOps/How Agentic AI can help for Cloud DevOps.md b/raw/Cloud & DevOps/How Agentic AI can help for Cloud DevOps.md similarity index 100% rename from raw/Technical/Cloud & DevOps/How Agentic AI can help for Cloud DevOps.md rename to raw/Cloud & DevOps/How Agentic AI can help for Cloud DevOps.md diff --git a/raw/Technical/Cloud & DevOps/How Can a Multi Cloud Strategy Transform Your Business ROI.md b/raw/Cloud & DevOps/How Can a Multi Cloud Strategy Transform Your Business ROI.md similarity index 100% rename from raw/Technical/Cloud & DevOps/How Can a Multi Cloud Strategy Transform Your Business ROI.md rename to raw/Cloud & DevOps/How Can a Multi Cloud Strategy Transform Your Business ROI.md diff --git a/raw/Technical/Cloud & DevOps/How to Simplify Multi-Account Deployments Monitoring Centralized Logs for AWS CloudFormation StackSets.md b/raw/Cloud & DevOps/How to Simplify Multi-Account Deployments Monitoring Centralized Logs for AWS CloudFormation StackSets.md similarity index 100% rename from raw/Technical/Cloud & DevOps/How to Simplify Multi-Account Deployments Monitoring Centralized Logs for AWS CloudFormation StackSets.md rename to raw/Cloud & DevOps/How to Simplify Multi-Account Deployments Monitoring Centralized Logs for AWS CloudFormation StackSets.md diff --git a/raw/Technical/Cloud & DevOps/Public vs Private vs Hybrid Cloud Differences Explained.md b/raw/Cloud & DevOps/Public vs Private vs Hybrid Cloud Differences Explained.md similarity index 100% rename from raw/Technical/Cloud & DevOps/Public vs Private vs Hybrid Cloud Differences Explained.md rename to raw/Cloud & DevOps/Public vs Private vs Hybrid Cloud Differences Explained.md diff --git a/raw/Technical/Cloud & DevOps/RTO vs RPO Key Differences for Modern Disaster Recovery.md b/raw/Cloud & DevOps/RTO vs RPO Key Differences for Modern Disaster Recovery.md similarity index 100% rename from raw/Technical/Cloud & DevOps/RTO vs RPO Key Differences for Modern Disaster Recovery.md rename to raw/Cloud & DevOps/RTO vs RPO Key Differences for Modern Disaster Recovery.md diff --git a/raw/Technical/Cloud & DevOps/The Myths and Misconceptions About Cloud Computing LinkedIn.md b/raw/Cloud & DevOps/The Myths and Misconceptions About Cloud Computing LinkedIn.md similarity index 100% rename from raw/Technical/Cloud & DevOps/The Myths and Misconceptions About Cloud Computing LinkedIn.md rename to raw/Cloud & DevOps/The Myths and Misconceptions About Cloud Computing LinkedIn.md diff --git a/raw/Technical/Cloud & DevOps/Understanding Complete ITSM.md b/raw/Cloud & DevOps/Understanding Complete ITSM.md similarity index 100% rename from raw/Technical/Cloud & DevOps/Understanding Complete ITSM.md rename to raw/Cloud & DevOps/Understanding Complete ITSM.md diff --git a/raw/Technical/Cloud & DevOps/What I know about Cloud Service Delivery 1.md b/raw/Cloud & DevOps/What I know about Cloud Service Delivery 1.md similarity index 100% rename from raw/Technical/Cloud & DevOps/What I know about Cloud Service Delivery 1.md rename to raw/Cloud & DevOps/What I know about Cloud Service Delivery 1.md diff --git a/raw/Technical/Cloud & DevOps/What is DevSecOps Best Practices, Benefits, and Tools.md b/raw/Cloud & DevOps/What is DevSecOps Best Practices, Benefits, and Tools.md similarity index 100% rename from raw/Technical/Cloud & DevOps/What is DevSecOps Best Practices, Benefits, and Tools.md rename to raw/Cloud & DevOps/What is DevSecOps Best Practices, Benefits, and Tools.md diff --git a/raw/Technical/Home Office/Install Apache Superset in Docker.md b/raw/Home Office/Install Apache Superset in Docker.md similarity index 100% rename from raw/Technical/Home Office/Install Apache Superset in Docker.md rename to raw/Home Office/Install Apache Superset in Docker.md diff --git a/raw/Technical/Home Office/Mac-Mini-服务器配置-防止自动锁屏与睡眠.md b/raw/Home Office/Mac-Mini-服务器配置-防止自动锁屏与睡眠.md similarity index 100% rename from raw/Technical/Home Office/Mac-Mini-服务器配置-防止自动锁屏与睡眠.md rename to raw/Home Office/Mac-Mini-服务器配置-防止自动锁屏与睡眠.md diff --git a/raw/Technical/Home Office/Synology NAS + Xiaoya Alist + CloudDrvie2+ Plex to Build Media Platform.md b/raw/Home Office/Synology NAS + Xiaoya Alist + CloudDrvie2+ Plex to Build Media Platform.md similarity index 100% rename from raw/Technical/Home Office/Synology NAS + Xiaoya Alist + CloudDrvie2+ Plex to Build Media Platform.md rename to raw/Home Office/Synology NAS + Xiaoya Alist + CloudDrvie2+ Plex to Build Media Platform.md diff --git a/raw/Technical/Home Office/可自动化、可扩展、AI增强的电商数据采集与处理系统.md b/raw/Home Office/可自动化、可扩展、AI增强的电商数据采集与处理系统.md similarity index 100% rename from raw/Technical/Home Office/可自动化、可扩展、AI增强的电商数据采集与处理系统.md rename to raw/Home Office/可自动化、可扩展、AI增强的电商数据采集与处理系统.md diff --git a/raw/Technical/Home Office/在 Ubuntu 安装 Ollama 并运行 Qwen2.5‑Coder 7B.md b/raw/Home Office/在 Ubuntu 安装 Ollama 并运行 Qwen2.5‑Coder 7B.md similarity index 100% rename from raw/Technical/Home Office/在 Ubuntu 安装 Ollama 并运行 Qwen2.5‑Coder 7B.md rename to raw/Home Office/在 Ubuntu 安装 Ollama 并运行 Qwen2.5‑Coder 7B.md diff --git a/raw/Technical/Home Office/通过VPS+内网反向代理实现域名访问内网穿透.md b/raw/Home Office/通过VPS+内网反向代理实现域名访问内网穿透.md similarity index 100% rename from raw/Technical/Home Office/通过VPS+内网反向代理实现域名访问内网穿透.md rename to raw/Home Office/通过VPS+内网反向代理实现域名访问内网穿透.md diff --git a/raw/Technical/Home Office/🟠3X-UI Xray on BandwagonVPS.md b/raw/Home Office/🟠3X-UI Xray on BandwagonVPS.md similarity index 100% rename from raw/Technical/Home Office/🟠3X-UI Xray on BandwagonVPS.md rename to raw/Home Office/🟠3X-UI Xray on BandwagonVPS.md diff --git a/raw/Technical/Home Office/🟠Clonezilla对Ubuntu Server进行全盘镜像备份.md b/raw/Home Office/🟠Clonezilla对Ubuntu Server进行全盘镜像备份.md similarity index 100% rename from raw/Technical/Home Office/🟠Clonezilla对Ubuntu Server进行全盘镜像备份.md rename to raw/Home Office/🟠Clonezilla对Ubuntu Server进行全盘镜像备份.md diff --git a/raw/Technical/Home Office/🟠Linux 运维必会的 150 个命令.md b/raw/Home Office/🟠Linux 运维必会的 150 个命令.md similarity index 100% rename from raw/Technical/Home Office/🟠Linux 运维必会的 150 个命令.md rename to raw/Home Office/🟠Linux 运维必会的 150 个命令.md diff --git a/raw/Technical/Home Office/🟠MinIO + Zipline 自托管图床应用安装教程.md b/raw/Home Office/🟠MinIO + Zipline 自托管图床应用安装教程.md similarity index 100% rename from raw/Technical/Home Office/🟠MinIO + Zipline 自托管图床应用安装教程.md rename to raw/Home Office/🟠MinIO + Zipline 自托管图床应用安装教程.md diff --git a/raw/Technical/Home Office/🟠MySQL MariaDB 数据库详细信息.md b/raw/Home Office/🟠MySQL MariaDB 数据库详细信息.md similarity index 100% rename from raw/Technical/Home Office/🟠MySQL MariaDB 数据库详细信息.md rename to raw/Home Office/🟠MySQL MariaDB 数据库详细信息.md diff --git a/raw/Technical/Home Office/🟠NodeWarden - 把 Bitwarden 搬上 Cloudflare Workers,彻底告别服务器.md b/raw/Home Office/🟠NodeWarden - 把 Bitwarden 搬上 Cloudflare Workers,彻底告别服务器.md similarity index 100% rename from raw/Technical/Home Office/🟠NodeWarden - 把 Bitwarden 搬上 Cloudflare Workers,彻底告别服务器.md rename to raw/Home Office/🟠NodeWarden - 把 Bitwarden 搬上 Cloudflare Workers,彻底告别服务器.md diff --git a/raw/Technical/Home Office/🟠RAX50 路由器 更新Merlin Clash订阅.md b/raw/Home Office/🟠RAX50 路由器 更新Merlin Clash订阅.md similarity index 100% rename from raw/Technical/Home Office/🟠RAX50 路由器 更新Merlin Clash订阅.md rename to raw/Home Office/🟠RAX50 路由器 更新Merlin Clash订阅.md diff --git a/raw/Technical/Home Office/🟠Ubuntu 24.04 enable SSH.md b/raw/Home Office/🟠Ubuntu 24.04 enable SSH.md similarity index 100% rename from raw/Technical/Home Office/🟠Ubuntu 24.04 enable SSH.md rename to raw/Home Office/🟠Ubuntu 24.04 enable SSH.md diff --git a/raw/Technical/Home Office/🟠Ubuntu Server科学上网.md b/raw/Home Office/🟠Ubuntu Server科学上网.md similarity index 100% rename from raw/Technical/Home Office/🟠Ubuntu Server科学上网.md rename to raw/Home Office/🟠Ubuntu Server科学上网.md diff --git a/raw/Technical/Home Office/🟠Ubuntu服务器通过rsync实现日常增量备份.md b/raw/Home Office/🟠Ubuntu服务器通过rsync实现日常增量备份.md similarity index 100% rename from raw/Technical/Home Office/🟠Ubuntu服务器通过rsync实现日常增量备份.md rename to raw/Home Office/🟠Ubuntu服务器通过rsync实现日常增量备份.md diff --git a/raw/Technical/Home Office/🟠Ubuntu用RustDesk远程登录出现不能使用Wayland登录的错误.md b/raw/Home Office/🟠Ubuntu用RustDesk远程登录出现不能使用Wayland登录的错误.md similarity index 100% rename from raw/Technical/Home Office/🟠Ubuntu用RustDesk远程登录出现不能使用Wayland登录的错误.md rename to raw/Home Office/🟠Ubuntu用RustDesk远程登录出现不能使用Wayland登录的错误.md diff --git a/raw/Technical/Home Office/🟠Ubuntu禁用合盖休眠.md b/raw/Home Office/🟠Ubuntu禁用合盖休眠.md similarity index 100% rename from raw/Technical/Home Office/🟠Ubuntu禁用合盖休眠.md rename to raw/Home Office/🟠Ubuntu禁用合盖休眠.md diff --git a/raw/Technical/Home Office/🟠在Synology NAS上安装CloudDrive2.md b/raw/Home Office/🟠在Synology NAS上安装CloudDrive2.md similarity index 100% rename from raw/Technical/Home Office/🟠在Synology NAS上安装CloudDrive2.md rename to raw/Home Office/🟠在Synology NAS上安装CloudDrive2.md diff --git a/raw/Technical/Home Office/🟠在Ubuntu上通过VPS+内网反向代理实现域名访问内网穿透.md b/raw/Home Office/🟠在Ubuntu上通过VPS+内网反向代理实现域名访问内网穿透.md similarity index 100% rename from raw/Technical/Home Office/🟠在Ubuntu上通过VPS+内网反向代理实现域名访问内网穿透.md rename to raw/Home Office/🟠在Ubuntu上通过VPS+内网反向代理实现域名访问内网穿透.md diff --git a/raw/Technical/Home Office/🟠如何传输Docker images 并且在另一个Docker安装.md b/raw/Home Office/🟠如何传输Docker images 并且在另一个Docker安装.md similarity index 100% rename from raw/Technical/Home Office/🟠如何传输Docker images 并且在另一个Docker安装.md rename to raw/Home Office/🟠如何传输Docker images 并且在另一个Docker安装.md diff --git a/raw/Technical/Home Office/🟠如何删除旧的废弃的docker container +volume.md b/raw/Home Office/🟠如何删除旧的废弃的docker container +volume.md similarity index 100% rename from raw/Technical/Home Office/🟠如何删除旧的废弃的docker container +volume.md rename to raw/Home Office/🟠如何删除旧的废弃的docker container +volume.md diff --git a/raw/Technical/Home Office/🟠如何判别你的Linux 服务器是 x64(也就是 x86_64)还是 ARM64.md b/raw/Home Office/🟠如何判别你的Linux 服务器是 x64(也就是 x86_64)还是 ARM64.md similarity index 100% rename from raw/Technical/Home Office/🟠如何判别你的Linux 服务器是 x64(也就是 x86_64)还是 ARM64.md rename to raw/Home Office/🟠如何判别你的Linux 服务器是 x64(也就是 x86_64)还是 ARM64.md diff --git a/raw/Technical/Home Office/🟠如何在Ubuntu Server上通过NFS挂载Synology NAS上的共享文件夹.md b/raw/Home Office/🟠如何在Ubuntu Server上通过NFS挂载Synology NAS上的共享文件夹.md similarity index 100% rename from raw/Technical/Home Office/🟠如何在Ubuntu Server上通过NFS挂载Synology NAS上的共享文件夹.md rename to raw/Home Office/🟠如何在Ubuntu Server上通过NFS挂载Synology NAS上的共享文件夹.md diff --git a/raw/Technical/Home Office/🟠如何在Ubuntu Server安装 docker & docker compose.md b/raw/Home Office/🟠如何在Ubuntu Server安装 docker & docker compose.md similarity index 100% rename from raw/Technical/Home Office/🟠如何在Ubuntu Server安装 docker & docker compose.md rename to raw/Home Office/🟠如何在Ubuntu Server安装 docker & docker compose.md diff --git a/raw/Technical/Home Office/🟠如何用指纹浏览器安全注册并订阅Claude Pro会员全攻略.md b/raw/Home Office/🟠如何用指纹浏览器安全注册并订阅Claude Pro会员全攻略.md similarity index 100% rename from raw/Technical/Home Office/🟠如何用指纹浏览器安全注册并订阅Claude Pro会员全攻略.md rename to raw/Home Office/🟠如何用指纹浏览器安全注册并订阅Claude Pro会员全攻略.md diff --git a/raw/Technical/Home Office/🟠安装Ubuntu-24.04.2在HP Zbook工作站笔记本上.md b/raw/Home Office/🟠安装Ubuntu-24.04.2在HP Zbook工作站笔记本上.md similarity index 100% rename from raw/Technical/Home Office/🟠安装Ubuntu-24.04.2在HP Zbook工作站笔记本上.md rename to raw/Home Office/🟠安装Ubuntu-24.04.2在HP Zbook工作站笔记本上.md diff --git a/raw/Technical/Home Office/🟠安装v2rayN.md b/raw/Home Office/🟠安装v2rayN.md similarity index 100% rename from raw/Technical/Home Office/🟠安装v2rayN.md rename to raw/Home Office/🟠安装v2rayN.md diff --git a/raw/Technical/Home Office/🟠家庭监控方案:Prometheus + Grafana + Node Exporter + cAdvisor +Blackbox.md b/raw/Home Office/🟠家庭监控方案:Prometheus + Grafana + Node Exporter + cAdvisor +Blackbox.md similarity index 100% rename from raw/Technical/Home Office/🟠家庭监控方案:Prometheus + Grafana + Node Exporter + cAdvisor +Blackbox.md rename to raw/Home Office/🟠家庭监控方案:Prometheus + Grafana + Node Exporter + cAdvisor +Blackbox.md diff --git a/raw/Technical/Home Office/🟠用Docker中安装Navidrome.md b/raw/Home Office/🟠用Docker中安装Navidrome.md similarity index 100% rename from raw/Technical/Home Office/🟠用Docker中安装Navidrome.md rename to raw/Home Office/🟠用Docker中安装Navidrome.md diff --git a/raw/Technical/Home Office/🟠用Docker安装Apache Superset.md b/raw/Home Office/🟠用Docker安装Apache Superset.md similarity index 100% rename from raw/Technical/Home Office/🟠用Docker安装Apache Superset.md rename to raw/Home Office/🟠用Docker安装Apache Superset.md diff --git a/raw/Technical/Home Office/🟠用Docker安装Homarr.md b/raw/Home Office/🟠用Docker安装Homarr.md similarity index 100% rename from raw/Technical/Home Office/🟠用Docker安装Homarr.md rename to raw/Home Office/🟠用Docker安装Homarr.md diff --git a/raw/Technical/Home Office/🟠用Docker安装Jellyfin.md b/raw/Home Office/🟠用Docker安装Jellyfin.md similarity index 100% rename from raw/Technical/Home Office/🟠用Docker安装Jellyfin.md rename to raw/Home Office/🟠用Docker安装Jellyfin.md diff --git a/raw/Technical/Home Office/🟠用Docker安装Portainer.md b/raw/Home Office/🟠用Docker安装Portainer.md similarity index 100% rename from raw/Technical/Home Office/🟠用Docker安装Portainer.md rename to raw/Home Office/🟠用Docker安装Portainer.md diff --git a/raw/Technical/Home Office/🟠用Docker安装it-tools.md b/raw/Home Office/🟠用Docker安装it-tools.md similarity index 100% rename from raw/Technical/Home Office/🟠用Docker安装it-tools.md rename to raw/Home Office/🟠用Docker安装it-tools.md diff --git a/raw/Technical/Home Office/🟠用Docker安装transmission.md b/raw/Home Office/🟠用Docker安装transmission.md similarity index 100% rename from raw/Technical/Home Office/🟠用Docker安装transmission.md rename to raw/Home Office/🟠用Docker安装transmission.md diff --git a/raw/Technical/Home Office/🟠网件RAX50路由器刷梅林固件与科学上网插件安装教程.md b/raw/Home Office/🟠网件RAX50路由器刷梅林固件与科学上网插件安装教程.md similarity index 100% rename from raw/Technical/Home Office/🟠网件RAX50路由器刷梅林固件与科学上网插件安装教程.md rename to raw/Home Office/🟠网件RAX50路由器刷梅林固件与科学上网插件安装教程.md diff --git a/raw/Technical/Home Office/🟠群晖NAS科学上网方法.md b/raw/Home Office/🟠群晖NAS科学上网方法.md similarity index 100% rename from raw/Technical/Home Office/🟠群晖NAS科学上网方法.md rename to raw/Home Office/🟠群晖NAS科学上网方法.md diff --git a/raw/Technical/Home Office/🟢家庭网络环境概览_2026-04-03.md b/raw/Home Office/🟢家庭网络环境概览_2026-04-03.md similarity index 100% rename from raw/Technical/Home Office/🟢家庭网络环境概览_2026-04-03.md rename to raw/Home Office/🟢家庭网络环境概览_2026-04-03.md diff --git a/raw/Technical/Home Office/🟣Mac Mini 安装 FRP 0.65.0(ARM64)操作笔记.md b/raw/Home Office/🟣Mac Mini 安装 FRP 0.65.0(ARM64)操作笔记.md similarity index 100% rename from raw/Technical/Home Office/🟣Mac Mini 安装 FRP 0.65.0(ARM64)操作笔记.md rename to raw/Home Office/🟣Mac Mini 安装 FRP 0.65.0(ARM64)操作笔记.md diff --git a/raw/Technical/Home Office/🟣Ubuntu 安装 FRP 0.65.0(x86_64)操作笔记.md b/raw/Home Office/🟣Ubuntu 安装 FRP 0.65.0(x86_64)操作笔记.md similarity index 95% rename from raw/Technical/Home Office/🟣Ubuntu 安装 FRP 0.65.0(x86_64)操作笔记.md rename to raw/Home Office/🟣Ubuntu 安装 FRP 0.65.0(x86_64)操作笔记.md index 026b9ac9..0a5fd7f5 100644 --- a/raw/Technical/Home Office/🟣Ubuntu 安装 FRP 0.65.0(x86_64)操作笔记.md +++ b/raw/Home Office/🟣Ubuntu 安装 FRP 0.65.0(x86_64)操作笔记.md @@ -420,5 +420,5 @@ sudo netstat -tlnp | grep <端口号> ## 十五、相关文档 -- [Mac Mini 安装 FRP 0.65.0(ARM64)操作笔记](./🟣Mac%20Mini%20安装%20FRP%200.65.0(ARM64)操作笔记.md) -- [通过VPS+内网反向代理实现域名访问内网穿透](./通过VPS+内网反向代理实现域名访问内网穿透.md) +- [Mac Mini 安装 FRP 0.65.0(ARM64)操作笔记](🟣Mac%20Mini%20安装%20FRP%200.65.0(ARM64)操作笔记.md) +- [通过VPS+内网反向代理实现域名访问内网穿透](通过VPS+内网反向代理实现域名访问内网穿透.md) diff --git a/raw/Technical/Home Office/🟣macOS 创建与解除 Symbolic Link(OpenClaw 目录映射).md b/raw/Home Office/🟣macOS 创建与解除 Symbolic Link(OpenClaw 目录映射).md similarity index 100% rename from raw/Technical/Home Office/🟣macOS 创建与解除 Symbolic Link(OpenClaw 目录映射).md rename to raw/Home Office/🟣macOS 创建与解除 Symbolic Link(OpenClaw 目录映射).md diff --git a/raw/Technical/Project/TikTok PM - Python Django Project.md b/raw/Project/TikTok PM - Python Django Project.md similarity index 100% rename from raw/Technical/Project/TikTok PM - Python Django Project.md rename to raw/Project/TikTok PM - Python Django Project.md diff --git a/raw/Technical/Vibe Coding/Vibe-Kanban + OpenCode 在 Ubuntu Server 上安装与管理指南.md b/raw/Vibe Coding/Vibe-Kanban + OpenCode 在 Ubuntu Server 上安装与管理指南.md similarity index 100% rename from raw/Technical/Vibe Coding/Vibe-Kanban + OpenCode 在 Ubuntu Server 上安装与管理指南.md rename to raw/Vibe Coding/Vibe-Kanban + OpenCode 在 Ubuntu Server 上安装与管理指南.md diff --git a/raw/Technical/Vibe Coding/vibe coding经验收集.md b/raw/Vibe Coding/vibe coding经验收集.md similarity index 100% rename from raw/Technical/Vibe Coding/vibe coding经验收集.md rename to raw/Vibe Coding/vibe coding经验收集.md diff --git a/raw/Technical/Vibe Coding/在Ubuntu上安装Vibe-Kanban.md b/raw/Vibe Coding/在Ubuntu上安装Vibe-Kanban.md similarity index 100% rename from raw/Technical/Vibe Coding/在Ubuntu上安装Vibe-Kanban.md rename to raw/Vibe Coding/在Ubuntu上安装Vibe-Kanban.md diff --git a/raw/Technical/Vibe Coding/如何在Ubuntu上安装opencode并配置Vibe-Kanban.md b/raw/Vibe Coding/如何在Ubuntu上安装opencode并配置Vibe-Kanban.md similarity index 100% rename from raw/Technical/Vibe Coding/如何在Ubuntu上安装opencode并配置Vibe-Kanban.md rename to raw/Vibe Coding/如何在Ubuntu上安装opencode并配置Vibe-Kanban.md diff --git a/raw/Technical/Vibe Coding/如何在项目里安装Claude-Code-Templates Skills.md b/raw/Vibe Coding/如何在项目里安装Claude-Code-Templates Skills.md similarity index 100% rename from raw/Technical/Vibe Coding/如何在项目里安装Claude-Code-Templates Skills.md rename to raw/Vibe Coding/如何在项目里安装Claude-Code-Templates Skills.md diff --git a/raw/Technical/Workflow/n8n configure telegram trigger.md b/raw/Workflow/n8n configure telegram trigger.md similarity index 100% rename from raw/Technical/Workflow/n8n configure telegram trigger.md rename to raw/Workflow/n8n configure telegram trigger.md diff --git a/raw/Technical/Workflow/n8n docker install & update.md b/raw/Workflow/n8n docker install & update.md similarity index 100% rename from raw/Technical/Workflow/n8n docker install & update.md rename to raw/Workflow/n8n docker install & update.md diff --git a/raw/Technical/Workflow/n8n full tutorial building AI agents in 2025 for Beginners!.md b/raw/Workflow/n8n full tutorial building AI agents in 2025 for Beginners!.md similarity index 100% rename from raw/Technical/Workflow/n8n full tutorial building AI agents in 2025 for Beginners!.md rename to raw/Workflow/n8n full tutorial building AI agents in 2025 for Beginners!.md diff --git a/raw/Technical/Workflow/n8n+Claude 通过自然语言自动化工作流.md b/raw/Workflow/n8n+Claude 通过自然语言自动化工作流.md similarity index 100% rename from raw/Technical/Workflow/n8n+Claude 通过自然语言自动化工作流.md rename to raw/Workflow/n8n+Claude 通过自然语言自动化工作流.md diff --git a/raw/Technical/Workflow/使用Claude自动生成N8N工作流的实操教程.md b/raw/Workflow/使用Claude自动生成N8N工作流的实操教程.md similarity index 100% rename from raw/Technical/Workflow/使用Claude自动生成N8N工作流的实操教程.md rename to raw/Workflow/使用Claude自动生成N8N工作流的实操教程.md diff --git a/跨境电商/Scrapy + Playwright 抓取TikTok Shop Data.md b/raw/跨境电商/Scrapy + Playwright 抓取TikTok Shop Data.md similarity index 100% rename from 跨境电商/Scrapy + Playwright 抓取TikTok Shop Data.md rename to raw/跨境电商/Scrapy + Playwright 抓取TikTok Shop Data.md diff --git a/跨境电商/TK美国面单授权及操作流程.md b/raw/跨境电商/TK美国面单授权及操作流程.md similarity index 100% rename from 跨境电商/TK美国面单授权及操作流程.md rename to raw/跨境电商/TK美国面单授权及操作流程.md diff --git a/跨境电商/TikTok Shop - Apache Superset Dashboard设计思路.md b/raw/跨境电商/TikTok Shop - Apache Superset Dashboard设计思路.md similarity index 100% rename from 跨境电商/TikTok Shop - Apache Superset Dashboard设计思路.md rename to raw/跨境电商/TikTok Shop - Apache Superset Dashboard设计思路.md diff --git a/跨境电商/做TK跨境思路不对努力白费.md b/raw/跨境电商/做TK跨境思路不对努力白费.md similarity index 100% rename from 跨境电商/做TK跨境思路不对努力白费.md rename to raw/跨境电商/做TK跨境思路不对努力白费.md diff --git a/跨境电商/电商如何选品 如何找到爆款 选品策略.md b/raw/跨境电商/电商如何选品 如何找到爆款 选品策略.md similarity index 100% rename from 跨境电商/电商如何选品 如何找到爆款 选品策略.md rename to raw/跨境电商/电商如何选品 如何找到爆款 选品策略.md diff --git a/跨境电商/电商视频Prompt.md b/raw/跨境电商/电商视频Prompt.md similarity index 100% rename from 跨境电商/电商视频Prompt.md rename to raw/跨境电商/电商视频Prompt.md diff --git a/跨境电商/超达物流定价.md b/raw/跨境电商/超达物流定价.md similarity index 100% rename from 跨境电商/超达物流定价.md rename to raw/跨境电商/超达物流定价.md diff --git a/wiki/concepts/AIOps.md b/wiki/concepts/AIOps.md new file mode 100644 index 00000000..91c72bd4 --- /dev/null +++ b/wiki/concepts/AIOps.md @@ -0,0 +1,25 @@ +--- +title: "AIOps" +type: concept +tags: [AI, Operations, Monitoring] +sources: [agentic-ai-cloud-devops.md, modern-itsm.md] +last_updated: 2025-03-01 +--- + +## Summary +AIOps是使用机器学习和AI自动化IT运维任务的实践。 + +## Definition +AIOps(AI for IT Operations)利用机器学习算法自动分析IT数据、检测异常、关联事件、预测问题。 + +## Key Capabilities +- 异常检测 +- 事件关联和分析 +- 根因分析 +- 预测性维护 +- 自我修复 + +## Connections +- [[Agentic AI]] ← 包含 ← AIOps +- [[ITSM]] ← 增强 ← AIOps + diff --git a/wiki/concepts/Agentic-AI.md b/wiki/concepts/Agentic-AI.md new file mode 100644 index 00000000..699db786 --- /dev/null +++ b/wiki/concepts/Agentic-AI.md @@ -0,0 +1,27 @@ +--- +title: "Agentic AI" +type: concept +tags: [AI, Automation, DevOps] +sources: [agentic-ai-cloud-devops.md] +last_updated: 2025-03-01 +--- + +## Summary +Agentic AI是具有自主决策和任务执行能力的AI系统,能够增强云DevOps能力。 + +## Definition +Agentic AI系统可以自主分析环境、做出决策并执行任务,无需持续人工干预。 + +## Application Domains +- 自主事件检测与解决 +- 自动化云部署和配置 +- 智能成本优化 +- AI驱动的安全与合规 +- 智能日志分析 +- 多租户管理自动化 +- AI增强决策 + +## Connections +- [[AIOps]] ← 包含 ← Agentic AI +- [[Self-Healing Systems]] ← 依赖 ← Agentic AI + diff --git a/wiki/concepts/Agentic-System.md b/wiki/concepts/Agentic-System.md new file mode 100644 index 00000000..7233a45c --- /dev/null +++ b/wiki/concepts/Agentic-System.md @@ -0,0 +1,23 @@ +--- +title: "Agentic System(代理系统)" +type: concept +tags: [ai, agent, n8n] +sources: [n8n-full-tutorial-building-ai-agents-in-2025-for-beginners.md] +last_updated: 2026-04-14 +--- + +## Definition +Agentic System(代理系统)是由 Agent 和 Workflow 组成的混合架构,结合工作流的确定性与 Agent 的灵活性,实现动态任务执行。 + +## Properties +- **组成**:Agent(基于 LLM 动态选择工具)+ Workflow(预定义自动化流程) +- **优势**:兼具可预测性和适应性 +- **应用场景**:客户服务、个性化服务、复杂任务自动化 + +## Related Concepts +- [[Agent模式]]:Agent 的自动执行能力 +- [[Memory Module]]:保留对话上下文的机制 + +## Connections +- [[n8n]] ← 实现 ← [[Agentic System]] +- [[Agentic System]] ← 使用 ← [[LLM]] \ No newline at end of file diff --git a/wiki/concepts/Agent模式.md b/wiki/concepts/Agent模式.md new file mode 100644 index 00000000..a583e561 --- /dev/null +++ b/wiki/concepts/Agent模式.md @@ -0,0 +1,24 @@ +--- +title: "Agent模式" +type: concept +tags: [AI, Cursor, 模式] +sources: [cursor-2-0-beginner-guide, mcp-in-cursor] +last_updated: 2026-04-14 +--- + +## Definition +Cursor中的交互模式,自动执行内嵌命令并处理工具调用。 + +## 与Normal模式对比 +| 模式 | 特点 | 风险 | +|------|------|------| +| Agent模式 | 自动执行命令 | 可能误操作 | +| Normal模式 | 需手动复制执行 | 较安全 | +| Ask模式 | 仅返回文本 | 不改动文件 | + +## YOLO Mode +- 开启后自动执行所有命令 +- 官方默认关闭,建议谨慎使用 + +## Connections +- [[Cursor]] ← 应用 ← [[Agent模式]] diff --git a/wiki/concepts/CI-CD.md b/wiki/concepts/CI-CD.md new file mode 100644 index 00000000..dced0d6c --- /dev/null +++ b/wiki/concepts/CI-CD.md @@ -0,0 +1,28 @@ +--- +title: "CI/CD" +type: concept +tags: [DevOps, Automation, Deployment] +sources: [devops-culture-transformation.md, devops-maturity-model.md, devsecops-best-practices.md] +last_updated: 2025-03-01 +--- + +## Summary +CI/CD是持续集成/持续交付的缩写,是现代软件开发的核心实践。 + +## Definition +CI/CD是自动化软件构建、测试和部署的实践,使小批量更新能够频繁可靠地发布。 + +## Components +- CI:持续集成,每次代码提交自动构建和测试 +- CD:持续交付,自动将通过测试的代码部署到各环境 +- 持续部署:自动将通过所有测试的代码部署到生产 + +## Key Claims +- CI/CD是Agile加速器,将反馈周期从数周缩短到数分钟 +- CI/CD是DevOps的核心使能器 +- DevSecOps将安全测试集成到CI/CD + +## Connections +- [[DevOps文化]] ← 实现 ← CI/CD +- [[DevSecOps]] ← 扩展 ← CI/CD + diff --git a/wiki/concepts/Caddy.md b/wiki/concepts/Caddy.md new file mode 100644 index 00000000..b34d0a2c --- /dev/null +++ b/wiki/concepts/Caddy.md @@ -0,0 +1,47 @@ +--- +title: "Caddy" +type: concept +tags: [caddy, reverse-proxy, https, web-server] +last_updated: 2026-04-13 +--- + +## Definition +Caddy是现代化Web服务器,自动提供HTTPS并支持反向代理。 + +## 核心特性 +- 自动Let's Encrypt证书申请 +- 自动HTTP→HTTPS重定向 +- 支持反向代理和负载均衡 +- 配置简洁 + +## 在家庭网络中的用途 +- 反向代理内网服务到公网域名 +- HTTPS终端处理 +- 多域名统一管理 + +## Caddyfile配置示例 +```caddy +n8n.ishenwei.online { + reverse_proxy 127.0.0.1:15678 +} + +nas.ishenwei.online { + reverse_proxy 127.0.0.1:15000 +} +``` + +## 常用命令 +```bash +# 验证配置 +sudo caddy validate --config /etc/caddy/Caddyfile + +# 重载配置 +sudo systemctl reload caddy + +# 重启服务 +sudo systemctl restart caddy +``` + +## 相关文档 +- [[通过VPS+内网反向代理实现域名访问内网穿透]] +- [[家庭网络环境概览 2026-04-03]] \ No newline at end of file diff --git a/wiki/concepts/Claude-Skills.md b/wiki/concepts/Claude-Skills.md new file mode 100644 index 00000000..a6b6adbe --- /dev/null +++ b/wiki/concepts/Claude-Skills.md @@ -0,0 +1,25 @@ +--- +title: "Claude Skills" +type: concept +tags: [AI技能, 工作流, 流程工程] +sources: [32wan-ren-shou-cang-de-claude-skills.md] +last_updated: 2026-04-14 +--- + +## Definition +写给AI的"说明书"和"SOP",将重复任务拆解为AI可理解、可复用、可自动执行的流程。 + +## Core Attributes +- 本质:说明书 + SOP +- 特征:可复用、可自动执行、有边界 +- 价值:从提示词工程升级到流程工程 + +## Related Concepts +- [[SOP]]:标准作业程序,Skills的核心组成 +- [[流程工程]]:Skills代表的AI应用新范式 +- [[Vibe Coding]] ← 尽头 ← [[Claude Skills]] + +## Tool Ecosystem +- 官方仓库:[[Claude Skills官方仓库]] +- 聚合站:[[skillsmp.com]], [[aitmpl.com]], [[claudemarketplaces.com]] +- 精选仓库:[[awesome-claude-skills]] diff --git a/wiki/concepts/Cloudflare-D1.md b/wiki/concepts/Cloudflare-D1.md new file mode 100644 index 00000000..ae00e6b0 --- /dev/null +++ b/wiki/concepts/Cloudflare-D1.md @@ -0,0 +1,21 @@ +--- +title: "Cloudflare D1" +type: concept +tags: [cloudflare, database, sqlite] +sources: [nodewarden-ba-bitwarden-ban-dao-cloudflare-workers-shang-che-di-gao-bie-fu-wu-qi.md] +last_updated: 2026-04-14 +--- + +## Definition +Cloudflare D1是Cloudflare提供的边缘SQLite数据库服务,基于Workers构建。 + +## Key Attributes +- 类型:边缘数据库 +- 引擎:SQLite +- 部署:Cloudflare全球网络 + +## Related Entities +- [[NodeWarden]]:使用D1存储密码元数据 + +## Connections +- [[Cloudflare D1]] ← 存储 ← [[NodeWarden]] diff --git a/wiki/concepts/Cloudflare-R2.md b/wiki/concepts/Cloudflare-R2.md new file mode 100644 index 00000000..12b16a51 --- /dev/null +++ b/wiki/concepts/Cloudflare-R2.md @@ -0,0 +1,21 @@ +--- +title: "Cloudflare R2" +type: concept +tags: [cloudflare, object-storage, s3] +sources: [nodewarden-ba-bitwarden-ban-dao-cloudflare-workers-shang-che-di-gao-bie-fu-wu-qi.md] +last_updated: 2026-04-14 +--- + +## Definition +Cloudflare R2是Cloudflare提供的对象存储服务,S3兼容且无出口流量费用。 + +## Key Attributes +- 类型:对象存储 +- 协议:S3兼容 +- 特色:无出口流量费用 + +## Related Entities +- [[NodeWarden]]:使用R2存储附件 + +## Connections +- [[Cloudflare R2]] ← 存储 ← [[NodeWarden]] diff --git a/wiki/concepts/Compaction.md b/wiki/concepts/Compaction.md new file mode 100644 index 00000000..6b0d4428 --- /dev/null +++ b/wiki/concepts/Compaction.md @@ -0,0 +1,35 @@ +--- +title: "Compaction(内存压缩)" +type: concept +tags: [AI, Agent, 上下文] +sources: [yang-xia-ri-ji-4, yang-long-xia-5-tian] +last_updated: 2026-04-14 +--- + +## Definition +OpenClaw在对话填满Context Window时,将旧消息压缩成摘要的机制。 + +## 问题 +- 摘要抓住要点,但丢失细节 +- 姓名、数字、具体决定统统消失 +- 所有消息一视同仁(重要指令和闲聊得到相同待遇) + +## 解决方案:Memory Flush +```json +{ + "compaction": { + "memoryFlush": { + "enabled": true, + "softThresholdTokens": 4000 + } + } +} +``` +在压缩运行前将重要上下文写入memory/YYYY-MM-DD.md + +## 关键洞察 +"压缩不是敌人。未写入的上下文才是。" — 比利哥 + +## Connections +- [[Context-Window]] ← 受限于 ← [[Compaction]] +- [[Memory-Flush]] ← 保护机制 ← [[Compaction]] diff --git a/wiki/concepts/Composer.md b/wiki/concepts/Composer.md new file mode 100644 index 00000000..9ca43cfe --- /dev/null +++ b/wiki/concepts/Composer.md @@ -0,0 +1,13 @@ +--- +title: "Composer" +type: concept +tags: [AI, Cursor, 模型] +sources: [cursor-2-0-beginner-guide] +last_updated: 2026-04-14 +--- + +## Definition +Cursor自研AI模型,主打生成速度比同类模型快4倍。 + +## Connections +- [[Cursor]] ← 核心模型 ← [[Composer]] diff --git a/wiki/concepts/Context-Window.md b/wiki/concepts/Context-Window.md new file mode 100644 index 00000000..0fcd69d6 --- /dev/null +++ b/wiki/concepts/Context-Window.md @@ -0,0 +1,42 @@ +--- +title: "Context Window" +type: concept +tags: [AI, 模型, Context] +sources: [yang-xia-ri-ji-4, yang-long-xia-5-tian] +last_updated: 2026-04-14 +--- + +## Definition +模型能处理的上下文大小,决定了对话长度和信息保留量。 + +## 典型值 +| 模型 | Context Window | +|------|---------------| +| MiniMax-M2.7 | 200K tokens | +| deepseek-reasoner | 16K tokens | + +## Compaction机制 +OpenClaw的对话压缩机制: +- 当对话填满Context Window时,将旧消息压缩成摘要 +- 摘要抓住要点,但丢失细节(姓名、数字、具体决定) +- safeguard模式预留一半token给compaction + +## Memory Flush解决方案 +压缩前将重要上下文写入磁盘: +```json +{ + "compaction": { + "memoryFlush": { + "enabled": true, + "softThresholdTokens": 4000 + } + } +} +``` + +## Connections +- [[Compaction]] ← 上下文压缩 ← [[Context-Window]] +- [[Memory-Flush]] ← 保护机制 ← [[Context-Window]] +- [[Embedding Vector]] ← 限制 ← [[Context-Window]] +- [[Generation]] ← 限制 ← [[Context-Window]] +- [[Indexing]] ← 限制 ← [[Context-Window]] diff --git a/wiki/concepts/DORA指标.md b/wiki/concepts/DORA指标.md new file mode 100644 index 00000000..af93aeff --- /dev/null +++ b/wiki/concepts/DORA指标.md @@ -0,0 +1,24 @@ +--- +title: "DORA指标" +type: concept +tags: [DevOps, Metrics] +sources: [cloud-devop-maturity-guideline.md, devops-maturity-model.md] +last_updated: 2025-03-01 +--- + +## Summary +DORA(DevOps Research and Assessment)指标是用于衡量组织DevOps性能和能力的四项关键指标。 + +## Definition +DORA指标由DORA研究组织提出,包含四项核心指标用于评估软件交付和组织绩效。 + +## Components +- 部署频率:代码部署到生产环境的频率 +- 变更前置时间:从代码提交到生产部署的时间 +- 变更失败率:导致生产失败或需要回滚的部署百分比 +- MTTR:Mean Time to Recovery,平均恢复时间 + +## Connections +- [[DevOps成熟度模型]] ← 依赖 ← DORA指标 +- [[MTTR]] ← 属于 ← DORA指标 + diff --git a/wiki/concepts/DRY原则.md b/wiki/concepts/DRY原则.md new file mode 100644 index 00000000..5fbaed64 --- /dev/null +++ b/wiki/concepts/DRY原则.md @@ -0,0 +1,18 @@ +--- +title: "DRY原则" +type: concept +tags: [编码, 原则] +sources: [dev-experience-standards] +last_updated: 2026-04-14 +--- + +## Definition +Don't Repeat Yourself,不重复代码原则。 + +## 实践 +- 提炼公共逻辑 +- 避免复制粘贴 +- 模块化复用 + +## Connections +- [[单一职责]] ← 关联 ← [[DRY原则]] diff --git a/wiki/concepts/DevOps成熟度模型.md b/wiki/concepts/DevOps成熟度模型.md new file mode 100644 index 00000000..9e8afa35 --- /dev/null +++ b/wiki/concepts/DevOps成熟度模型.md @@ -0,0 +1,31 @@ +--- +title: "DevOps成熟度模型" +type: concept +tags: [DevOps, Maturity] +sources: [cloud-devop-maturity-guideline.md, devops-maturity-model.md] +last_updated: 2025-03-01 +--- + +## Summary +DevOps成熟度模型是评估组织DevOps实践阶段性发展的框架,帮助组织理解当前位置并规划提升路径。 + +## Definition +DevOps成熟度模型是一种结构化框架,通过多个阶段评估组织的DevOps实践水平,从初始的临时阶段到完全优化的成熟阶段。 + +## Key Claims +- DevOps成熟度模型帮助组织分析衡量当前DevOps能力和方法论 +- 可建立基准、定义目标成熟度级别、识别需改进的关键领域 +- DORA指标是DevOps成熟度评估的核心指标集 +- DevOps是持续改进过程,即使成熟公司也需适应新技术 + +## Dimensions +- 文化与战略 +- 自动化 +- 结构与流程 +- 协作与共享 +- 技术 + +## Connections +- [[DORA指标]] ← 衡量 ← DevOps成熟度 +- [[DevSecOps]] ← 扩展 ← DevOps成熟度模型 + diff --git a/wiki/concepts/DevSecOps.md b/wiki/concepts/DevSecOps.md new file mode 100644 index 00000000..00729ea2 --- /dev/null +++ b/wiki/concepts/DevSecOps.md @@ -0,0 +1,29 @@ +--- +title: "DevSecOps" +type: concept +tags: [DevOps, Security, SDLC] +sources: [devsecops-best-practices.md, devops-culture-transformation.md, devops-maturity-model.md] +last_updated: 2025-03-01 +--- + +## Summary +DevSecOps是将安全集成到DevOps生命周期每个阶段的实践,确保安全是共同责任。 + +## Definition +DevSecOps = Development + Security + Operations,将安全实践嵌入整个软件开发生命周期。 + +## Key Claims +- 70%发布后发现的安全漏洞可通过DevSecOps预防 +- 安全不是最后检查,而是从编码开始就集成 +- Shift Left:将安全缺陷识别提前到开发早期 + +## Core Components +- 协作:开发、安全、运维团队共享安全责任 +- 自动化:将安全测试集成到CI/CD流水线 +- 测试:SAST、DAST、SCA、IAST + +## Connections +- [[DevOps成熟度模型]] ← 包含 ← DevSecOps +- [[SDLC]] ← 集成 ← DevSecOps +- [[CI/CD]] ← 包含 ← 安全测试 + diff --git a/wiki/concepts/Django-Admin定制.md b/wiki/concepts/Django-Admin定制.md new file mode 100644 index 00000000..59cd067f --- /dev/null +++ b/wiki/concepts/Django-Admin定制.md @@ -0,0 +1,59 @@ +--- +title: "Django Admin定制" +type: concept +tags: [django, admin, customization] +sources: [tiktok-pm-python-django-project.md] +last_updated: 2026-04-14 +--- + +## Definition +Django Admin是Django框架自带的管理后台,可通过配置实现复杂的CRUD操作界面。 + +## Key Customization Points + +### list_display +定义列表页显示的字段: +```python +list_display = ('source_id', 'title_short', 'store_name', 'final_price', 'sold') +``` + +### search_fields +配置快速关键词搜索: +```python +search_fields = ('source_id', 'title', 'store_name', 'category', 'seller_id') +``` + +### list_filter +配置多条件过滤侧边栏: +```python +list_filter = ('store_name', 'category', 'currency', 'final_price') +``` + +### fieldsets +字段分组显示: +```python +fieldsets = ( + ('Product Base Info', { + 'fields': (('source_id', 'title'),) + }), +) +``` + +### inlines +内联关联模型(图片、视频、变体、评价): +```python +inlines = [ProductVariationInline, ProductImageInline, ProductVideoInline, ProductReviewInline] +``` + +### readonly_fields +不可编辑字段: +```python +readonly_fields = ('source_id', 'created_at', 'updated_at') +``` + +## Image Preview Modal +通过自定义CSS和JavaScript实现点击图片放大功能。 + +## Connections +- [[Django Admin定制]] ← part_of ← [[Django]] +- [[Django Admin定制]] ← manages ← [[Django ORM模型设计]] diff --git a/wiki/concepts/Docker代理配置.md b/wiki/concepts/Docker代理配置.md new file mode 100644 index 00000000..7e1e539b --- /dev/null +++ b/wiki/concepts/Docker代理配置.md @@ -0,0 +1,22 @@ +--- +title: "Docker代理配置" +type: concept +tags: [docker, proxy, container] +sources: [ubuntu-server-ke-xue-shang-wang.md] +last_updated: 2026-04-14 +--- + +## Definition +Docker代理配置解决Docker Daemon层和容器层需要单独配置代理的问题。 + +## Key Attributes +- Daemon代理:修改systemd service环境变量 +- 容器代理:修改~/.docker/config.json或运行时-e参数 +- 关键点:容器内使用宿主机网桥IP而非127.0.0.1 + +## Related Entities +- [[Docker]]:需要配置代理 +- [[V2Ray/Tuic]]:代理来源 + +## Connections +- [[Docker代理配置]] ← 代理源 ← [[V2Ray/Tuic]] diff --git a/wiki/concepts/Docker容器化部署.md b/wiki/concepts/Docker容器化部署.md new file mode 100644 index 00000000..96146049 --- /dev/null +++ b/wiki/concepts/Docker容器化部署.md @@ -0,0 +1,57 @@ +--- +title: "Docker容器化部署" +type: concept +tags: [docker, deployment, containerization] +sources: [tiktok-pm-python-django-project.md] +last_updated: 2026-04-14 +--- + +## Definition +Docker容器化部署通过docker-compose编排多个容器,实现开发生产环境一致性。 + +## Architecture + +### Services +- **web**:Django + Gunicorn应用 +- **nginx**:反向代理和静态文件服务 +- **MySQL**:外部数据库(也可容器化) + +### docker-compose.yml +```yaml +services: + web: + build: . + command: gunicorn tiktok_pm_project.wsgi:application + volumes: + - ./data:/app/data + env_file: + - .env + + worker: + build: . + command: python manage.py qcluster + + nginx: + image: nginx + ports: + - "80:80" + volumes: + - ./nginx/nginx.conf:/etc/nginx/nginx.conf + depends_on: + - web +``` + +## Deployment Workflow +1. 拉取最新代码:git pull origin main +2. 重建容器:docker compose up --build -d +3. 执行迁移:docker compose exec web python manage.py migrate +4. 创建超级用户:docker compose exec web python manage.py createsuperuser + +## Static Files +使用collectstatic收集静态文件,Nginx提供/static/路由。 + +## Connections +- [[Docker容器化部署]] ← contains ← [[Django]] +- [[Docker容器化部署]] ← contains ← [[Gunicorn]] +- [[Docker容器化部署]] ← contains ← [[Nginx]] +- [[Docker容器化部署]] ← uses ← [[MySQL]] diff --git a/wiki/concepts/Embedding-Vector.md b/wiki/concepts/Embedding-Vector.md new file mode 100644 index 00000000..5db8a8ec --- /dev/null +++ b/wiki/concepts/Embedding-Vector.md @@ -0,0 +1,23 @@ +--- +title: "Embedding Vector" +type: concept +tags: [Embedding, 向量, 语义表示] +--- + +## Definition +Embedding Vector是文本的数值表示,捕获语义信息,将文本转为固定长度向量。 + +## Core Mechanism +- 通过深度学习模型将文本编码为稠密向量 +- 语义相似的文本在向量空间中距离相近 +- 通常128-2048维浮点数向量 + +## Key Properties +- 受Context Window限制(单次编码的文本长度) +- 依赖Embedding Model质量 +- 计算相似度通常用余弦相似度或点积 + +## Connections +- [[Indexing]] ← 输出 ← [[Embedding Vector]] +- [[Retrieval]] ← 依赖 ← [[Embedding Vector]] +- [[Vector Store]] ← 存储 ← [[Embedding Vector]] diff --git a/wiki/concepts/Extended-Thinking.md b/wiki/concepts/Extended-Thinking.md new file mode 100644 index 00000000..bf5dba99 --- /dev/null +++ b/wiki/concepts/Extended-Thinking.md @@ -0,0 +1,21 @@ +--- +title: "Extended Thinking(扩展思考)" +type: concept +tags: [claude, AI, reasoning] +sources: [n8n-Claude-通过自然语言自动化工作流.md, 使用Claude自动生成N8N工作流的实操教程.md] +last_updated: 2026-04-14 +--- + +## Definition +Claude 的一种运行模式,支持更深层次的逻辑推理,使 AI 在生成复杂任务(如工作流代码)时表现更好。 + +## Usage +在 Claude 配置中开启 Extended Thinking 模式,可提升代码生成质量。 + +## Related Concepts +- [[Opensea模型]]:与 Extended Thinking 配合使用的 Claude 子模型 +- [[自然语言工作流生成]]:Extended Thinking 提升此场景效果 + +## Connections +- [[Claude]] ← 使用 ← [[Extended Thinking]] +- [[自然语言工作流生成]] ← 增强 ← [[Extended Thinking]] \ No newline at end of file diff --git a/wiki/concepts/FRP.md b/wiki/concepts/FRP.md new file mode 100644 index 00000000..c3d4c898 --- /dev/null +++ b/wiki/concepts/FRP.md @@ -0,0 +1,52 @@ +--- +title: "FRP" +type: concept +tags: [frp, 内网穿透, reverse-proxy] +last_updated: 2026-04-13 +--- + +## Definition +FRP(Fast Reverse Proxy)是一款高性能内网穿透工具,支持TCP/UDP/HTTP/HTTPS协议。 + +## 组件 +| 组件 | 说明 | 运行位置 | +|------|------|----------| +| frps | 服务端 | VPS | +| frpc | 客户端 | 内网机器 | + +## 核心配置 +### frps.ini (VPS) +```ini +[common] +bind_addr = 0.0.0.0 +bind_port = 7000 +dashboard_addr = 0.0.0.0 +dashboard_port = 7500 +token = your_token_here +``` + +### frpc.toml (内网客户端) +```toml +[common] +server_addr = VPS_IP +server_port = 7000 +token = your_token_here + +[[proxies]] +name = "ssh" +type = "tcp" +localIP = "127.0.0.1" +localPort = 22 +remotePort = 60022 +``` + +## 版本支持 +| 架构 | 版本 | 安装路径 | +|------|------|----------| +| macOS ARM64 | 0.65.0 | /opt/frp/frp_0.65.0_darwin_arm64 | +| Linux amd64 | 0.65.0 | /opt/frp/frp_0.65.0_linux_amd64 | + +## 相关文档 +- [[Mac Mini 安装 FRP 0.65.0(ARM64)操作笔记]] +- [[Ubuntu 安装 FRP 0.65.0(x86_64)操作笔记]] +- [[通过VPS+内网反向代理实现域名访问内网穿透]] \ No newline at end of file diff --git a/wiki/concepts/FinOps.md b/wiki/concepts/FinOps.md new file mode 100644 index 00000000..82bd7e1c --- /dev/null +++ b/wiki/concepts/FinOps.md @@ -0,0 +1,24 @@ +--- +title: "FinOps" +type: concept +tags: [Cloud, Cost Optimization, Financial] +sources: [cloud-operating-model.md, multi-cloud-strategy-roi.md, cloud-service-delivery.md] +last_updated: 2025-03-01 +--- + +## Summary +FinOps(云财务运营)是一种云成本管理实践,平衡云支出与业务价值。 + +## Definition +FinOps是云财务管理框架,通过实时监控、优化和问责实现云成本效率。 + +## Key Practices +- 实时成本监控 +- 资源优化和Right-sizing +- Reserved Instances和Savings Plans利用 +- 成本可见性和报告 + +## Connections +- [[云运营模型]] ← 包含 ← FinOps +- [[多云策略]] ← 受益 ← FinOps + diff --git a/wiki/concepts/Gemini-3应用开发方法论.md b/wiki/concepts/Gemini-3应用开发方法论.md new file mode 100644 index 00000000..41a1eb97 --- /dev/null +++ b/wiki/concepts/Gemini-3应用开发方法论.md @@ -0,0 +1,24 @@ +--- +id: Gemini-3应用开发方法论 +title: "Gemini-3应用开发方法论" +type: concept +tags: [gemini, application-development, prompt-engineering] +sources: [] +last_updated: 2025-12-18 +--- + +# Gemini-3应用开发方法论 + +三步法快速构建AI应用的工作流程。 + +## Definition +利用Gemini-3等大模型,通过约束输入、扩展思考、设计输出的方式快速构建AI应用。 + +## Steps +1. **思考输入的场景**:局限输入词汇在垂直场景(如诗词、小说、电影等) +2. **约束模型的思考**:利用提示词、MCP将输入词汇扩展为结构化内容 +3. **设计输出的容器**:使用前端代码(SVG/HTML)可视化模型输出内容 + +## Related Concepts +- [[结构化输出]] +- [[提示词优化]] \ No newline at end of file diff --git a/wiki/concepts/Generation.md b/wiki/concepts/Generation.md new file mode 100644 index 00000000..6f9e9768 --- /dev/null +++ b/wiki/concepts/Generation.md @@ -0,0 +1,25 @@ +--- +title: "Generation" +type: concept +tags: [RAG, LLM, 生成] +--- + +## Definition +Generation是将问题+知识片段输入LLM生成答案的过程,是RAG的第三阶段。 + +## Core Mechanism +1. 知识组装:将检索到的文档片段与问题组合 +2. Prompt构建:构造包含上下文的Prompt +3. LLM生成:调用LLM生成答案 +4. 输出后处理:答案格式化/过滤 + +## Key Properties +- 受Context Window限制 +- 依赖LLM推理能力 +- 检索质量决定生成质量 + +## Connections +- [[RAG]] ← 阶段3 ← [[Generation]] +- [[Retrieval]] ← 上游 ← [[Generation]] +- [[LLM]] ← 依赖 ← [[Generation]] +- [[Context Window]] ← 限制 ← [[Generation]] diff --git a/wiki/concepts/Git代理配置.md b/wiki/concepts/Git代理配置.md new file mode 100644 index 00000000..0ab0e898 --- /dev/null +++ b/wiki/concepts/Git代理配置.md @@ -0,0 +1,31 @@ +--- +title: "Git代理配置" +type: concept +tags: [Git, 代理, GitHub] +sources: [git-push-connection-reset] +last_updated: 2026-04-14 +--- + +## Definition +为Git单独设置HTTP或SOCKS5代理,解决国内访问GitHub的连接问题。 + +## HTTP代理配置 +```bash +git config --global http.proxy http://127.0.0.1:10809 +git config --global https.proxy http://127.0.0.1:10809 +``` + +## SOCKS5代理配置 +```bash +git config --global http.proxy socks5://127.0.0.1:10808 +git config --global https.proxy socks5://127.0.0.1:10808 +``` + +## 取消代理 +```bash +git config --global --unset http.proxy +git config --global --unset https.proxy +``` + +## Connections +- [[GitHub]] ← 解决方案 ← [[Git代理配置]] diff --git a/wiki/concepts/ITSM.md b/wiki/concepts/ITSM.md new file mode 100644 index 00000000..fd5f83ae --- /dev/null +++ b/wiki/concepts/ITSM.md @@ -0,0 +1,34 @@ +--- +title: "ITSM" +type: concept +tags: [IT Service Management, Operations] +sources: [modern-itsm.md] +last_updated: 2025-03-01 +--- + +## Summary +ITSM(IT服务管理)是管理IT服务交付、运营和持续改进的战略性方法。 + +## Definition +ITSM是用于设计、交付、管理和改进IT服务的框架和实践集合。 + +## Modern Trends +- AI驱动的异常检测和预测分析 +- 实时可观测性和自动化修复 +- AIOps驱动的自我修复IT生态系统 +- Policy-as-Code合规自动化 +- 零信任安全架构 + +## Key Domains +- 问题管理 +- 事件管理 +- 变更管理 +- 配置管理 +- 资产管理 +- 安全与合规 +- 灾难恢复 + +## Connections +- [[AIOps]] ← 驱动 ← 现代ITSM +- [[零信任架构]] ← 集成 ← ITSM + diff --git a/wiki/concepts/Indexing.md b/wiki/concepts/Indexing.md new file mode 100644 index 00000000..73baca4b --- /dev/null +++ b/wiki/concepts/Indexing.md @@ -0,0 +1,25 @@ +--- +title: "Indexing" +type: concept +tags: [RAG, 索引, 数据处理] +--- + +## Definition +Indexing是将外部文档切分并建立索引的过程,是RAG的第一阶段。 + +## Core Mechanism +1. 文档加载(Load) +2. 文档切分(Split):按段落、句子或Token切分 +3. 向量化(Embed):通过Embedding Model转为向量 +4. 存储(Store):存入Vector Store + +## Key Properties +- 切分策略影响检索质量 +- 受Context Window限制 +- 需要平衡粒度 + +## Connections +- [[RAG]] ← 阶段1 ← [[Indexing]] +- [[Retrieval]] ← 下游 ← [[Indexing]] +- [[Embedding Vector]] ← 输出 ← [[Indexing]] +- [[Vector Store]] ← 目标 ← [[Indexing]] diff --git a/wiki/concepts/LLM-Wiki.md b/wiki/concepts/LLM-Wiki.md new file mode 100644 index 00000000..d8a93a4f --- /dev/null +++ b/wiki/concepts/LLM-Wiki.md @@ -0,0 +1,32 @@ +--- +title: "LLM Wiki" +type: concept +tags: [AI, 知识管理, RAG] +sources: [yang-xia-ri-ji-3] +last_updated: 2026-04-14 +--- + +## Definition +AI增量构建和维护持久化Wiki的系统,页面之间互相链接,知识越积越厚。与RAG模式形成对比。 + +## RAG vs LLM Wiki + +| 维度 | RAG | LLM Wiki | +|------|-----|----------| +| 检索方式 | 每次从零检索 | 增量累积 | +| 知识积累 | 不积累 | 越积越厚 | +| 页面关联 | 无 | 双向链接 | +| 维护方式 | 被动等着被查询 | 主动维护和更新 | + +## 核心洞察(Karpathy) +RAG模式是"每次从零检索",知识不积累;而LLM Wiki是让AI增量构建和维护一个持久化的Wiki。 + +## 在实践中的体现 +- AI执行任务过程中顺手维护链接、更新摘要、添加Tag、标记新旧矛盾 +- 不需要主动维护文档,AI在执行任务过程中顺手更新 +- Graph View发现孤岛页面和知识盲区 + +## Connections +- [[RAG]] ← 对比 → [[LLM-Wiki]] +- [[Obsidian]] ← 工具载体 ← [[LLM-Wiki]] +- [[Karpathy]] ← 提出 ← [[LLM-Wiki]] diff --git a/wiki/concepts/MCP.md b/wiki/concepts/MCP.md new file mode 100644 index 00000000..a328de1c --- /dev/null +++ b/wiki/concepts/MCP.md @@ -0,0 +1,24 @@ +--- +title: "MCP" +type: concept +tags: [AI, 协议, 集成] +sources: [mcp-in-cursor] +last_updated: 2026-04-14 +--- + +## Definition +Modal Context Protocol,基于Client-Server架构的协议,用于AI大模型与外围工具的集成交互。 + +## 核心接口 +| 接口 | 类型 | 功能 | +|------|------|------| +| 资源获取 | GET | 获取外部资源 | +| 工具调用 | POST | 执行外部工具 | +| Promise提示词 | - | 多样化交互 | + +## Cursor接入方式 +- **SSE方式**:通过Server-Sent Events接入 +- **Command方式**:通过本地执行命令接入 + +## Connections +- [[Cursor]] ← 集成 ← [[MCP]] diff --git a/wiki/concepts/Memory-Flush.md b/wiki/concepts/Memory-Flush.md new file mode 100644 index 00000000..b1c81eb2 --- /dev/null +++ b/wiki/concepts/Memory-Flush.md @@ -0,0 +1,40 @@ +--- +title: "Memory Flush" +type: concept +tags: [AI, Agent, 记忆] +sources: [yang-long-xia-5-tian] +last_updated: 2026-04-14 +--- + +## Definition +OpenClaw在对话压缩前将重要上下文写入磁盘的机制,防止压缩过程中丢失重要信息。 + +## 工作原理 +```json +{ + "compaction": { + "memoryFlush": { + "enabled": true, + "softThresholdTokens": 4000 + } + } +} +``` + +当会话接近上下文限制(4000 tokens)时: +1. 触发静默回合 +2. 提醒Agent将重要事实写入memory/YYYY-MM-DD.md +3. 压缩运行 +4. 即使上下文摘要丢失,重要内容仍在磁盘上 + +## 注意 +4000这个数值要根据模型context window调整: +- 大模型(32K/128K/200K)用4000太保守,会导致频繁summary +- 小模型可以用更激进的配置 + +## 关键洞察 +"压缩不是敌人。压缩过程中丢失信息才是。" — 比利哥 + +## Connections +- [[Compaction]] ← 配合 ← [[Memory-Flush]] +- [[Context-Window]] ← 阈值依据 ← [[Memory-Flush]] diff --git a/wiki/concepts/Opensea模型.md b/wiki/concepts/Opensea模型.md new file mode 100644 index 00000000..597e537e --- /dev/null +++ b/wiki/concepts/Opensea模型.md @@ -0,0 +1,21 @@ +--- +title: "Opensea模型" +type: concept +tags: [claude, AI, code-generation] +sources: [使用Claude自动生成N8N工作流的实操教程.md] +last_updated: 2026-04-14 +--- + +## Definition +Claude 的一个子模型,专为代码生成任务优化,适合自动编程场景。 + +## Usage +使用 Claude 生成 N8N 工作流时,推荐切换到 Opensea 模型以获得更好的代码生成效果。 + +## Related Concepts +- [[Extended Thinking]]:与 Opensea 配合使用效果更佳 +- [[自然语言工作流生成]]:Opensea 专为此场景优化 + +## Connections +- [[Claude]] ← 使用 ← [[Opensea模型]] +- [[自然语言工作流生成]] ← 使用 ← [[Opensea模型]] \ No newline at end of file diff --git a/wiki/concepts/RESTful-API设计.md b/wiki/concepts/RESTful-API设计.md new file mode 100644 index 00000000..cf81d223 --- /dev/null +++ b/wiki/concepts/RESTful-API设计.md @@ -0,0 +1,55 @@ +--- +title: "RESTful API设计" +type: concept +tags: [api, rest, django-rest-framework] +sources: [tiktok-pm-python-django-project.md] +last_updated: 2026-04-14 +--- + +## Definition +RESTful API是一种基于HTTP协议的API设计风格,使用URL表示资源,使用HTTP方法表示操作。 + +## Implementation in TikTok PM + +### ViewSet +使用Django REST Framework的ModelViewSet自动生成CRUD路由: +```python +class ProductViewSet(viewsets.ModelViewSet): + queryset = Product.objects.all() + serializer_class = ProductSerializer +``` + +### Router +使用DefaultRouter自动生成路由: +```python +router = DefaultRouter() +router.register(r'products', views.ProductViewSet) +``` + +### Serializers +序列化器将Django模型转换为JSON: +```python +class ProductSerializer(serializers.ModelSerializer): + images = ProductImageSerializer(many=True, read_only=True) + class Meta: + model = Product + exclude = ['raw_json', 'input'] +``` + +### Filtering +启用django-filter进行多条件过滤: +```python +filter_backends = [DjangoFilterBackend, filters.SearchFilter] +filterset_fields = ['available', 'category', 'seller_id', 'final_price'] +``` + +### API Routes +- GET /api/products/:查询所有产品 +- GET /api/products/?search=关键词:关键词搜索 +- GET /api/products/?category=服饰&In_stock=true:多条件过滤 +- POST /api/products/:创建产品 + +## Connections +- [[RESTful API设计]] ← uses ← [[Django]] +- [[RESTful API设计]] ← integrated_with ← [[Django Admin定制]] +- [[RESTful API设计]] ← consumed_by ← [[n8n]] diff --git a/wiki/concepts/RPO.md b/wiki/concepts/RPO.md new file mode 100644 index 00000000..0124f23e --- /dev/null +++ b/wiki/concepts/RPO.md @@ -0,0 +1,23 @@ +--- +title: "RPO" +type: concept +tags: [Disaster Recovery, Metrics] +sources: [rto-vs-rpo-disaster-recovery.md] +last_updated: 2025-07-26 +--- + +## Summary +RPO(恢复点目标)是组织可接受的最大数据丢失量指标。 + +## Definition +RPO是从故障点向前衡量的可接受数据丢失时间范围,例如上次备份是下午2点而数据库下午3点崩溃,则RPO为1小时。 + +## Key Claims +- RPO是关于数据的:可接受丢失多少最近数据 +- 备份频率高不一定有帮助,如果恢复时间长 +- 快速功能切换可保持低RPO同时不丢失数据 + +## Connections +- [[RTO]] ← 互补 ← RPO +- [[灾难恢复]] ← 依赖 ← RPO + diff --git a/wiki/concepts/RTO.md b/wiki/concepts/RTO.md new file mode 100644 index 00000000..e763f29d --- /dev/null +++ b/wiki/concepts/RTO.md @@ -0,0 +1,29 @@ +--- +title: "RTO" +type: concept +tags: [Disaster Recovery, Metrics] +sources: [rto-vs-rpo-disaster-recovery.md] +last_updated: 2025-07-26 +--- + +## Summary +RTO(恢复时间目标)是衡量系统可容忍最大停机时间的关键指标。 + +## Definition +RTO是系统可容忍的最大停机时间,从系统故障到恢复正常运行的时间目标。 + +## Key Claims +- RTO是关于速度的:系统能多快恢复在线 +- 功能标志可将RTO从数小时缩短到秒级 +- 不同应用层级应有不同的RTO目标 + +## Tiers +- Critical: < 5分钟 +- Important: < 1小时 +- Nice-to-have: < 4小时 + +## Connections +- [[RPO]] ← 互补 ← RTO +- [[功能标志]] ← 改善 ← RTO +- [[灾难恢复]] ← 依赖 ← RTO + diff --git a/wiki/concepts/Retrieval.md b/wiki/concepts/Retrieval.md new file mode 100644 index 00000000..a8d9fd83 --- /dev/null +++ b/wiki/concepts/Retrieval.md @@ -0,0 +1,25 @@ +--- +title: "Retrieval" +type: concept +tags: [RAG, 检索, 向量搜索] +--- + +## Definition +Retrieval是根据问题语义向量检索相关文档的过程,是RAG的第二阶段。 + +## Core Mechanism +1. 问题向量化:将用户问题通过Embedding Model转为向量 +2. 相似度计算:在Vector Store中计算相似度 +3. Top-K选取:返回最相关的K个文档片段 + +## Key Properties +- 依赖语义相似度 +- 受Embedding Model质量影响 +- 可结合关键词过滤 + +## Connections +- [[RAG]] ← 阶段2 ← [[Retrieval]] +- [[Indexing]] ← 上游 ← [[Retrieval]] +- [[Generation]] ← 下游 ← [[Retrieval]] +- [[Embedding Vector]] ← 依赖 ← [[Retrieval]] +- [[Vector Store]] ← 数据源 ← [[Retrieval]] diff --git a/wiki/concepts/S3兼容存储.md b/wiki/concepts/S3兼容存储.md new file mode 100644 index 00000000..51cfad4c --- /dev/null +++ b/wiki/concepts/S3兼容存储.md @@ -0,0 +1,19 @@ +--- +title: "S3兼容存储" +type: concept +tags: [object-storage, s3, cloud] +sources: [minio-zipline-zi-cheng-guan-tu-chuang-ying-yong-an-zhuang-jiao-cheng.md] +last_updated: 2026-04-14 +--- + +## Definition +S3兼容存储是支持Amazon S3 API的对象存储系统,可在本地或私有云部署。 + +## Key Attributes +- 协议:S3 REST API +- 访问控制:BUCKET/对象级别权限 +- 一致性:最终一致性 + +## Related Entities +- [[MinIO]]:S3兼容存储的实现 +- [[Zipline]]:使用S3存储的图片 \ No newline at end of file diff --git a/wiki/concepts/SSH协议切换.md b/wiki/concepts/SSH协议切换.md new file mode 100644 index 00000000..21eeec9b --- /dev/null +++ b/wiki/concepts/SSH协议切换.md @@ -0,0 +1,25 @@ +--- +title: "SSH协议切换" +type: concept +tags: [Git, SSH, GitHub] +sources: [git-push-connection-reset] +last_updated: 2026-04-14 +--- + +## Definition +从HTTPS切换到SSH协议连接GitHub,绕过443端口限制。 + +## 操作步骤 +```bash +# 查看当前远程地址 +git remote -v + +# 修改为SSH地址 +git remote set-url origin git@github.com:username/repo.git +``` + +## 前提 +已生成SSH Key并上传公钥到GitHub + +## Connections +- [[GitHub]] ← 解决方案 ← [[SSH协议切换]] diff --git a/wiki/concepts/Self-Improving-Skill.md b/wiki/concepts/Self-Improving-Skill.md new file mode 100644 index 00000000..79f27e98 --- /dev/null +++ b/wiki/concepts/Self-Improving-Skill.md @@ -0,0 +1,49 @@ +--- +title: "Self-Improving Skill" +type: concept +tags: [AI, Agent, 记忆, 复盘] +sources: [yang-xia-ri-ji-2] +last_updated: 2026-04-14 +--- + +## Definition +OpenClaw的结构化经验记录系统,每当Agent遇到问题、做出决策或发现值得记住的东西时,将内容写入LEARNINGS.md或ERRORS.md。 + +## 记录格式 +```markdown +## [LRN-YYYYMMDD-NNN] type + +**Logged**: YYYY-MM-DDTHH:MM:SS+08:00 +**Priority**: high/medium/low +**Status**: pending/resolved +**Area**: config/workflow/error + +### Summary +一句话描述学到了什么 + +### Details +具体发生了什么、问题出在哪 + +### Suggested Action +以后遇到类似情况该怎么做 + +### Metadata +- Pattern-Key: cron.telegram-delivery +- Recurrence-Count: 1 +- See Also: LRN-YYYYMMDD-NNN +``` + +## 核心价值 +- **错误只犯一次**:Pattern-Key重复是系统性问题的信号 +- **固定格式**:让人和其他Agent能快速检索、对比、追踪完整生命周期 + +## Pattern-Key使用 +| Pattern-Key | 出现次数 | 含义 | +|-------------|---------|------| +| cron.daily-self-review | 9次 | 每日复盘相关(活跃优化区)| +| cron.telegram-delivery | 2次 | Telegram通知配置(第二次解决)| + +## Connections +- [[三层记忆架构]] ← 自改进层 ← [[Self-Improving Skill]] +- [[Pattern-Key]] ← 分类机制 ← [[Self-Improving Skill]] +- [[Recurrence-Count]] ← 重复指标 ← [[Self-Improving Skill]] diff --git a/wiki/concepts/Skill蒸馏.md b/wiki/concepts/Skill蒸馏.md new file mode 100644 index 00000000..ca217734 --- /dev/null +++ b/wiki/concepts/Skill蒸馏.md @@ -0,0 +1,38 @@ +--- +title: "Skill蒸馏" +type: concept +tags: [AI, Skill, 框架] +sources: [yang-xia-ri-ji-5] +last_updated: 2026-04-14 +--- + +## Definition +通过女娲·Skill框架,从大量公开信息中提炼真实人物的核心思维框架,构建可运行的AI Skill。 + +## 蒸馏vs造人 +"女娲造人"是比喻: +- 不是从虚无中创造角色 +- 而是从大量信息中蒸馏核心心智模型 + +## 产出 +- 3-7个核心心智模型 +- 5-10条决策启发式 +- 一套表达DNA(高频用词、幽默风格、表达习惯) +- 价值观与诚实边界 + +## 采集维度(以苏东坡为例) +| Agent | 维度 | 采集内容 | +|-------|------|---------| +| 1 | 著作 | 诗词文章、《东坡易传》《东坡志林》 | +| 2 | 对话 | 与弟弟苏辙的书信、与佛印的对话 | +| 3 | 表达DNA | 高频用词、自嘲式幽默、蜀地方言痕迹 | +| 4 | 他者视角 | 林语堂《苏东坡传》、余秋雨《苏东坡突围》 | +| 5 | 决策 | 乌台诗案、黄州躬耕、惠州插秧 | +| 6 | 时间线 | 从眉山出生到北归病逝常州 | + +## 核心价值 +"每个人的Skill都是一个认知操作系统。你不需要同意他的所有观点,但你可以在需要的时候,用他的镜片看自己的问题。" — 女娲·Skill + +## Connections +- [[女娲·Skill]] ← 框架 ← [[Skill蒸馏]] +- [[认知操作系统]] ← 本质 ← [[Skill蒸馏]] diff --git a/wiki/concepts/Socket激活机制.md b/wiki/concepts/Socket激活机制.md new file mode 100644 index 00000000..b02a4e02 --- /dev/null +++ b/wiki/concepts/Socket激活机制.md @@ -0,0 +1,22 @@ +--- +title: "Socket激活机制" +type: concept +tags: [systemd, ssh, ubuntu] +sources: [ubuntu-24-04-enable-ssh.md] +last_updated: 2026-04-14 +--- + +## Definition +Socket激活机制是systemd的一种按需启动服务方式,服务在首次连接请求进入时才启动进程。 + +## Key Attributes +- 机制:有连接请求时才启动服务 +- 适用:Ubuntu 24.04默认SSH使用ssh.socket +- 对比:传统ssh.service为常驻运行 + +## Related Entities +- [[OpenSSH]]:可配置为socket激活 +- [[Ubuntu-Server]]:24.04版本引入了此机制 + +## Connections +- [[Socket激活机制]] ← 应用 ← [[OpenSSH]] diff --git a/wiki/concepts/Symbolic-Link.md b/wiki/concepts/Symbolic-Link.md new file mode 100644 index 00000000..35f08807 --- /dev/null +++ b/wiki/concepts/Symbolic-Link.md @@ -0,0 +1,44 @@ +--- +title: "Symbolic Link" +type: concept +tags: [symbolic-link, unix, macos, filesystem] +sources: [macos-chuang-jian-yu-jie-chu-symbolic-link-openclaw-mu-lu-ying-she] +last_updated: 2026-04-14 +--- + +## Definition +Symbolic Link(符号链接)是Unix/Linux系统中指向另一个文件的特殊文件类型,相当于Windows的快捷方式。 + +## 核心特性 +- 只删除链接文件,不删除真实目录 +- 支持目录和文件链接 +- 跨文件系统链接 +- 相对路径和绝对路径支持 + +## macOS创建命令 +```bash +# 创建符号链接 +ln -s /Users/weishen/.openclaw /Users/weishen/openclaw + +# 验证链接 +ls -l ~ | grep openclaw + +# 查看链接指向 +readlink ~/openclaw +``` + +## 解除链接 +```bash +rm ~/openclaw # 只删除链接,不删除真实目录 +``` + +## 注意事项 +- 不要误删真实目录:`rm -rf ~/.openclaw`会删除OpenClaw数据目录 +- 推荐使用推荐目录结构便于Git管理和备份 + +## OpenClaw应用场景 +将~/.openclaw(隐藏目录)映射到~/openclaw(普通目录),使Obsidian可直接访问。 + +## Connections +- [[OpenClaw]] ← 目录映射 ← [[Symbolic Link]] +- [[Obsidian]] ← 兼容 ← [[Symbolic Link]] diff --git a/wiki/concepts/TUI.md b/wiki/concepts/TUI.md new file mode 100644 index 00000000..07904724 --- /dev/null +++ b/wiki/concepts/TUI.md @@ -0,0 +1,26 @@ +--- +title: "TUI" +type: concept +tags: [Linux, 终端, 界面] +sources: [linux-system-monitor] +last_updated: 2026-04-14 +--- + +## Definition +Text User Interface,文本用户界面,在终端中运行的图形化界面。 + +## 优势 +- 响应快,资源占用低 +- SSH远程友好 +- 即使GUI卡顿时也能正常使用 + +## 工具对比 +| 工具 | 特点 | +|------|------| +| btop++ | 最佳TUI监控,美观与功能平衡 | +| htop | 聚焦进程,轻量 | +| Glances | 更轻量,完全键盘驱动 | +| Bottom | 专注实时图表 | + +## Connections +- [[btop++]] ← 类型 ← [[TUI]] diff --git a/wiki/concepts/Vector-Store.md b/wiki/concepts/Vector-Store.md new file mode 100644 index 00000000..b6e2f41a --- /dev/null +++ b/wiki/concepts/Vector-Store.md @@ -0,0 +1,23 @@ +--- +title: "Vector Store" +type: concept +tags: [向量数据库, 存储, RAG] +--- + +## Definition +Vector Store是存储Embedding Vector并实现相似度检索的数据库。 + +## Core Mechanism +- 高维向量存储 +- ANN(近似最近邻)搜索算法 +- 支持过滤条件和分桶 + +## Key Properties +- 支持海量向量高效检索 +- 常见实现:Qdrant、Milvus、Pinecone、Chroma +- 可与传统数据库结合 + +## Connections +- [[Embedding Vector]] ← 存储 ← [[Vector Store]] +- [[Retrieval]] ← 数据源 ← [[Vector Store]] +- [[Indexing]] ← 目标 ← [[Vector Store]] diff --git a/wiki/concepts/Vibe-Coding.md b/wiki/concepts/Vibe-Coding.md new file mode 100644 index 00000000..4f269b02 --- /dev/null +++ b/wiki/concepts/Vibe-Coding.md @@ -0,0 +1,36 @@ +--- +id: Vibe-Coding +title: "Vibe Coding" +type: concept +tags: [ai-agent, programming-paradigm] +aliases: [Vibe-Kanban] +--- + +## Definition +Vibe Coding 是一种 AI 辅助编程范式,通过自然语言描述需求,AI Agent 自动生成代码。其核心理念是"需求 → 伪代码 → 代码"的流程,以及"验证代码按正确逻辑运行"而非"看懂代码"。 + +## Core Properties +- **类型**:编程范式 +- **核心流程**:需求 → 伪代码 → 代码 +- **迭代方式**:点 → 线 → 体的逐级迭代 + +## Key Practices +1. **伪代码前置**:将设计文档写得很细,包括 Service 层的具体逻辑都用伪代码描述 +2. **双 AI 审查**:一个 AI 生成代码,另一个 AI review 后再修改 +3. **逐级迭代**:先打磨单个基础任务,再基于此批量执行 +4. **文件头注释**:描述代码作用、上下游链路,降低认知负载 +5. **自动化验证**:通过自动化测试、静态分析、形式化验证确保行为正确 + +## Related Entities +- [[Vibe-Kanban]]:Vibe Coding 的任务管理工具 +- [[OpenCode]]:支持 Vibe Coding 的 AI 编码 Agent +- [[Claude-Code-Templates]]:提供 Vibe Coding 相关模板 + +## Related Concepts +- [[Agent模式]]:AI Agent 自动执行命令的工作模式 +- [[TUI]]:终端用户界面,OpenCode 的主要交互形式 + +## Connections +- [[Vibe-Coding]] ← 使用 ← [[Vibe-Kanban]] +- [[Vibe-Coding]] ← 使用 ← [[OpenCode]] +- [[Vibe-Coding]] ← 依赖 ← [[Agent模式]] diff --git a/wiki/concepts/n8n-mcp.md b/wiki/concepts/n8n-mcp.md new file mode 100644 index 00000000..6dcce0ed --- /dev/null +++ b/wiki/concepts/n8n-mcp.md @@ -0,0 +1,35 @@ +--- +title: "n8n-mcp" +type: concept +tags: [n8n, MCP, AI] +sources: [n8n-Claude-通过自然语言自动化工作流.md, 使用Claude自动生成N8N工作流的实操教程.md] +last_updated: 2026-04-14 +--- + +## Definition +n8n 多功能控制面板(Multi-Controll Panel),作为 n8n 与 AI 模型之间的桥梁,使 AI 能够理解和使用 n8n 节点。 + +## Properties +- **节点覆盖**:543 个 n8n 节点(来自 n8n-nodes-base 和 @n8n/n8n-nodes-langchain) +- **属性覆盖**:99% +- **操作覆盖**:63.6% 的可用动作 +- **文档覆盖**:87%(包括 AI 节点) +- **AI 节点**:271 个具有完整文档的 AI 能力节点 +- **模板库**:2,709 个工作流模板,100% 元数据覆盖 +- **实际示例**:2,646 个从流行模板预提取的配置 + +## Mechanism +通过 n8n-mcp,Claude 可以: +1. 访问所有 n8n 节点定义 +2. 获取节点属性和操作 +3. 查看文档和示例 +4. 生成工作流配置 + +## Related Concepts +- [[MCP]]:Model Context Protocol,此处的技术基础 +- [[自然语言工作流生成]]:n8n-mcp 使此功能成为可能 + +## Connections +- [[n8n]] ← 暴露 ← [[n8n-mcp]] +- [[Claude]] ← 使用 ← [[n8n-mcp]] +- [[n8n-mcp]] ← 基于 ← [[Node.js]] \ No newline at end of file diff --git a/wiki/concepts/n8n节点类型.md b/wiki/concepts/n8n节点类型.md new file mode 100644 index 00000000..b3d31eaf --- /dev/null +++ b/wiki/concepts/n8n节点类型.md @@ -0,0 +1,22 @@ +--- +title: "n8n节点类型" +type: concept +tags: [n8n, workflow] +sources: [n8n-full-tutorial-building-ai-agents-in-2025-for-beginners.md] +last_updated: 2026-04-14 +--- + +## Definition +n8n 平台中节点的 5 种分类,每种类型承担不同功能: + +1. **触发器(Trigger)**:启动工作流的事件源 +2. **动作(Action)**:执行具体操作的节点 +3. **工具(Utility)**:数据处理转换节点 +4. **代码(Code)**:自定义 JavaScript/Python 代码节点 +5. **AI Agent**:高级 AI 能力节点,支持 LLM 调用 + +## Usage +理解节点分类是设计有效 n8n 工作流的基础。 + +## Connections +- [[n8n]] ← 使用 ← [[n8n节点类型]] \ No newline at end of file diff --git a/wiki/concepts/三层记忆架构.md b/wiki/concepts/三层记忆架构.md new file mode 100644 index 00000000..63352e4a --- /dev/null +++ b/wiki/concepts/三层记忆架构.md @@ -0,0 +1,30 @@ +--- +title: "三层记忆架构" +type: concept +tags: [AI, Agent, 记忆] +sources: [yang-xia-ri-ji-2] +last_updated: 2026-04-14 +--- + +## Definition +OpenClaw Agent的短期记忆 + 长期记忆 + self-improving复盘的三层记忆系统。 + +## 三层结构 + +| 层级 | 载体 | 功能 | 特点 | +|------|------|------|------| +| 短期记忆 | memory/YYYY-MM-DD.md | 每日对话记录 | 原始、仅追加 | +| 长期记忆 | LanceDB向量数据库 | 重要决策、偏好、流程 | 语义搜索 | +| 自改进层 | LEARNINGS.md | 每日23:00复盘、错误记录 | Pattern-Key追踪 | + +## 各司其职 +- **每日文件管上下文**:Agent启动时读取,接上昨天工作 +- **向量数据库管知识**:语义搜索找回重要信息 +- **self-improving管成长**:错误只犯一次,持续进化 + +## 核心观点 +"三层各司其职:每日文件管上下文,向量数据库管知识,self-improving管成长。" — 比利哥 + +## Connections +- [[Self-Improving Skill]] ← 自改进层 ← [[三层记忆架构]] +- [[LanceDB]] ← 长期记忆 ← [[三层记忆架构]] diff --git a/wiki/concepts/云成熟度模型.md b/wiki/concepts/云成熟度模型.md new file mode 100644 index 00000000..b2f1a0bd --- /dev/null +++ b/wiki/concepts/云成熟度模型.md @@ -0,0 +1,31 @@ +--- +title: "云成熟度模型" +type: concept +tags: [Cloud, Maturity] +sources: [cloud-maturity-model-guide.md] +last_updated: 2025-02-28 +--- + +## Summary +云成熟度模型(CMM)帮助组织评估云采用准备度,从业务和技术维度制定云迁移策略。 + +## Definition +云成熟度模型是评估组织云采用能力的阶段性框架,涵盖0-5级成熟度水平。 + +## Levels +- Level 0: 无云准备(遗留) +- Level 1: 初始准备(临时) +- Level 2: 可重复机会主义 +- Level 3: 系统化文档化 +- Level 4: 可测量 +- Level 5: 优化 + +## Key Dimensions +- 人员:技能和培训需求 +- 流程:云迁移和运营流程 +- 技术:云基础设施架构 + +## Connections +- [[云运营模型]] ← 依赖 ← 云成熟度评估 +- [[FinOps]] ← 成熟度支撑 ← 云成熟度模型 + diff --git a/wiki/concepts/云盘挂载.md b/wiki/concepts/云盘挂载.md new file mode 100644 index 00000000..b765d9e3 --- /dev/null +++ b/wiki/concepts/云盘挂载.md @@ -0,0 +1,36 @@ +--- +title: "云盘挂载" +type: concept +tags: [cloud-drive, aliyun, nfs, storage] +last_updated: 2025-02-23 +--- + +## Definition +云盘挂载是将云存储服务(如阿里云盘)映射为本地文件系统的技术。 + +## 工具 +| 工具 | 说明 | +|------|------| +| CloudDrive2 | 多云盘挂载工具 | +| Xiaoya Alist | 资源聚合和转存 | + +## 在家庭网络中的用途 +- 将阿里云盘作为本地硬盘使用 +- 通过Plex/Jellyfin直接播放云盘内容 +- 利用NAS做本地缓存加速 + +## CloudDrive2配置 +1. 安装矿神源后安装CloudDrive2 +2. 执行:sudo sed -i 's/package/root/g' /var/packages/CloudDrive2/conf/privilege +3. 访问:http://NAS_IP:19798/ +4. 阿里云盘App扫码授权(仅授权资源目录) + +## Xiaoya Alist配置 +1. 准备三个token文件: + - myopentoken.txt:开放token + - mytoken.txt:阿里云盘token + - temp_transfer_folder_id.txt:转存目录ID +2. Docker启动xiaoya/alist + +## 相关文档 +- [[Synology NAS + Xiaoya Alist + CloudDrive2 + Plex 家庭影视平台]] \ No newline at end of file diff --git a/wiki/concepts/云运营模型.md b/wiki/concepts/云运营模型.md new file mode 100644 index 00000000..dd568112 --- /dev/null +++ b/wiki/concepts/云运营模型.md @@ -0,0 +1,25 @@ +--- +title: "云运营模型" +type: concept +tags: [Cloud, Operating Model, Governance] +sources: [cloud-operating-model.md] +last_updated: 2025-03-01 +--- + +## Summary +云运营模型(COM)是标准化组织如何管理云资源、安全、自动化和成本的框架。 + +## Definition +云运营模型是确保云投资得到有效、安全、可持续管理的结构化框架。 + +## Four Pillars +1. 治理与合规:策略、标准、角色和决策流程 +2. 自动化与编排:IaC、DevOps工作流 +3. 安全与风险管理:零信任、加密、监控 +4. 云财务管理(FinOps):成本追踪、优化、预算控制 + +## Connections +- [[云成熟度模型]] ← 支撑 ← 云运营模型 +- [[FinOps]] ← 属于 ← 云运营模型支柱 +- [[Zero Trust安全模型]] ← 属于 ← 安全支柱 + diff --git a/wiki/concepts/交接协议.md b/wiki/concepts/交接协议.md new file mode 100644 index 00000000..1cbe2326 --- /dev/null +++ b/wiki/concepts/交接协议.md @@ -0,0 +1,29 @@ +--- +title: "交接协议" +type: concept +tags: [AI, Agent, 记忆] +sources: [yang-long-xia-5-tian] +last_updated: 2026-04-14 +--- + +## Definition +在模型切换或会话结束前,Agent将当前上下文写入每日日志的协议,解决模型切换时上下文丢失问题。 + +## 问题背景 +OpenClaw Agent在切换模型时: +- 新模型以新鲜上下文窗口开始 +- 只看到自动加载的文件 +- 不知道切换前发生了什么 + +## 解决方案 +在任何模型切换或会话结束前: +1. Agent将当前上下文写入memory/YYYY-MM-DD.md +2. 新会话启动时读取该文件 +3. 接上之前的工作 + +## 关键洞察 +"交接协议是模型切换的修复。" — 比利哥 + +## Connections +- [[写入纪律]] ← 执行机制 ← [[交接协议]] +- [[Context-Window]] ← 问题根源 ← [[交接协议]] diff --git a/wiki/concepts/内网穿透.md b/wiki/concepts/内网穿透.md new file mode 100644 index 00000000..952f831b --- /dev/null +++ b/wiki/concepts/内网穿透.md @@ -0,0 +1,37 @@ +--- +title: "内网穿透" +type: concept +tags: [frp, network, proxy,穿透] +last_updated: 2026-04-13 +--- + +## Definition +内网穿透是一种通过公网服务器中转,使外部网络能够访问内网服务的技术。 + +## 核心机制 +1. **FRP(Fast Reverse Proxy)**:高性能内网穿透工具 + - frps:服务端,运行在VPS上 + - frpc:客户端,运行在内网机器上 + +## 架构 +``` +公网客户端 → VPS(frps) → 内网机器(frpc) → 内网服务 +``` + +## 常见应用场景 +- 远程SSH访问内网服务器 +- 通过域名访问内网Web服务 +- 暴露内网端口到公网 + +## 在家庭网络中的应用 +| 工具 | 用途 | +|------|------| +| FRP | TCP/HTTP隧道 | +| Caddy | HTTPS反向代理 | +| Cloudflare | DNS托管 | + +## 相关文档 +- [[Mac Mini 安装 FRP 0.65.0(ARM64)操作笔记]] +- [[Ubuntu 安装 FRP 0.65.0(x86_64)操作笔记]] +- [[通过VPS+内网反向代理实现域名访问内网穿透]] +- [[家庭网络环境概览 2026-04-03]] \ No newline at end of file diff --git a/wiki/concepts/写入纪律.md b/wiki/concepts/写入纪律.md new file mode 100644 index 00000000..4bc60bbb --- /dev/null +++ b/wiki/concepts/写入纪律.md @@ -0,0 +1,39 @@ +--- +title: "写入纪律" +type: concept +tags: [AI, Agent, 记忆] +sources: [yang-long-xia-5-tian] +last_updated: 2026-04-14 +--- + +## Definition +强制要求Agent将任务结果、决策、错误记录到磁盘的纪律,是临时上下文变成永久记忆的方式。 + +## 核心观点 +"写入纪律比读取纪律更重要。" — 比利哥 + +## 为什么重要 +- 临时上下文存在于对话窗口中 +- 对话窗口会被压缩 +- 写盘的内容才能在压缩中存活 + +## 实践要点 +- 任务结果必须记录到磁盘 +- 每个错误变成一条规则写入LEARNINGS.md +- 交接协议:在模型切换前将当前上下文写入每日日志 + +## 自动加载文件(7个) +只有这些文件自动加载: +1. AGENTS.md +2. SOUL.md +3. TOOLS.md +4. IDENTITY.md +5. USER.md +6. HEARTBEAT.md +7. MEMORY.md + +其他一切需要明确读取指令。 + +## Connections +- [[交接协议]] ← 模型切换保护 ← [[写入纪律]] +- [[LEARNINGS.md]] ← 写入目标 ← [[写入纪律]] diff --git a/wiki/concepts/功能标志.md b/wiki/concepts/功能标志.md new file mode 100644 index 00000000..f1091c1b --- /dev/null +++ b/wiki/concepts/功能标志.md @@ -0,0 +1,25 @@ +--- +title: "功能标志" +type: concept +tags: [Deployment, Feature Flags, DevOps] +sources: [rto-vs-rpo-disaster-recovery.md] +last_updated: 2025-07-26 +--- + +## Summary +功能标志是控制功能开启/关闭的开关,无需重新部署即可改变系统行为。 + +## Definition +功能标志允许团队在不重新部署代码的情况下切换功能状态,支持渐进式发布和即时回滚。 + +## Key Claims +- 部署与发布解耦:可随时部署,择机发布 +- 渐进式发布:1%→5%→25%→100% +- Kill Switch:紧急情况下立即关闭问题功能 +- 将回滚时间从数小时缩短到秒级 + +## Connections +- [[RTO]] ← 改善 ← 功能标志 +- [[RPO]] ← 保护 ← 功能标志 +- [[LaunchDarkly]] ← 平台 ← 功能标志 + diff --git a/wiki/concepts/单一职责.md b/wiki/concepts/单一职责.md new file mode 100644 index 00000000..bde94554 --- /dev/null +++ b/wiki/concepts/单一职责.md @@ -0,0 +1,26 @@ +--- +title: "单一职责" +type: concept +tags: [编码, 原则] +sources: [dev-experience-standards] +last_updated: 2026-04-14 +--- + +## Definition +每个文件、类、函数只负责一件事的软件设计原则。 + +## 实践 +- 提炼公共逻辑 +- 避免重复代码(DRY) +- 模块化、函数化,提高复用价值 + +## 输入输出划分 +| 概念 | 说明 | +|------|------| +| 消费端 | 接收外部数据或依赖输入 | +| 生产端 | 生成数据、输出结果 | +| 状态 | 存储当前系统信息的变量 | +| 变换 | 处理状态、改变数据的逻辑 | + +## Connections +- [[开发经验与规范]] ← 原则 ← [[单一职责]] diff --git a/wiki/concepts/变量命名规范.md b/wiki/concepts/变量命名规范.md new file mode 100644 index 00000000..26d034ca --- /dev/null +++ b/wiki/concepts/变量命名规范.md @@ -0,0 +1,23 @@ +--- +title: "变量命名规范" +type: concept +tags: [编码, 规范] +sources: [dev-experience-standards] +last_updated: 2026-04-14 +--- + +## Definition +软件开发中的变量命名标准,确保代码可读性和可维护性。 + +## 规范 +- 格式:小写英文+下划线 或 小驼峰 +- 原则:命名体现内容职责,避免缩写 +- 常量:大写+下划线(如 `MAX_RETRY_COUNT`) +- 禁止:无意义名称如 `a, b, c` + +## 变量名大全文件 +建立统一索引文件,包含: +| 变量名 | 变量注释 | 出现位置 | 出现频率 | + +## Connections +- [[开发经验与规范]] ← 组成部分 ← [[变量命名规范]] diff --git a/wiki/concepts/品味值链.md b/wiki/concepts/品味值链.md new file mode 100644 index 00000000..97dc54f2 --- /dev/null +++ b/wiki/concepts/品味值链.md @@ -0,0 +1,26 @@ +--- +title: "品味值链" +type: concept +tags: [AI, 职业, 思维] +sources: [bu-tan-ji-shu] +last_updated: 2026-04-14 +--- + +## Definition +AI工具民主化后,品味成为差异化竞争力的概念。 + +## 背景 +- 1984年Mac发布,突然每个人都能做桌面出版 +- 90%的人做出来的东西丑得要命 +- 工具民主化了,但品味没有民主化 + +## AI时代同样 +- 现在每个人都能用AI写代码、做设计、生成内容 +- 90%的人用AI生成的东西是shit +- 因为他们不知道什么是好的 + +## 品味即护城河 +"你能判断AI给你的10个方案里哪个是insanely great的,你就比那些只会点'生成'按钮的人强一百倍。" — 乔布斯.skill + +## Connections +- [[乔布斯.skill]] ← 来源 ← [[品味值链]] diff --git a/wiki/concepts/固定点理论.md b/wiki/concepts/固定点理论.md new file mode 100644 index 00000000..61054e8e --- /dev/null +++ b/wiki/concepts/固定点理论.md @@ -0,0 +1,20 @@ +--- +title: "固定点理论" +type: concept +tags: [数学, 理论计算机科学, 递归] +sources: [formalization-recursive-self-optimizing-generative-systems.md] +last_updated: 2026-04-14 +--- + +## Definition +数学中,固定点是经过函数映射后保持不变的点。在递归自优化系统中,稳定生成能力对应于Φ的固定点。 + +## Core Attributes +- 定义:G* ∈ G,Φ(G*) = G* +- 含义:生成器在其自身的生成-优化-更新循环下保持不变 +- 收敛条件:Φ满足适当连续性或收缩性条件时,G* = lim(n→∞) Φ^n(G₀) + +## Related Concepts +- [[递归自优化]] ← 应用 ← [[固定点理论]] +- [[λ演算]] ← 表达工具 ← [[固定点理论]] +- [[Y组合子]] ← 实现机制 ← [[固定点理论]] diff --git a/wiki/concepts/图片托管服务.md b/wiki/concepts/图片托管服务.md new file mode 100644 index 00000000..53990f0f --- /dev/null +++ b/wiki/concepts/图片托管服务.md @@ -0,0 +1,23 @@ +--- +title: "图片托管服务" +type: concept +tags: [image-hosting, web-service] +sources: [minio-zipline-zi-cheng-guan-tu-chuang-ying-yong-an-zhuang-jiao-cheng.md] +last_updated: 2026-04-14 +--- + +## Definition +图片托管服务是提供图片上传、存储和访问URL生成的专用服务平台。 + +## Key Attributes +- 存储:对象存储(S3兼容) +- 访问:公开URL或签名URL +- 认证:API Token + +## Related Entities +- [[Zipline]]:自托管图片托管服务实现 +- [[MinIO]]:存储后端 + +## Connections +- [[图片托管服务]] ← 实现 ← [[Zipline]] +- [[图片托管服务]] ← 存储 ← [[S3兼容存储]] diff --git a/wiki/concepts/图生视频.md b/wiki/concepts/图生视频.md new file mode 100644 index 00000000..c30e586d --- /dev/null +++ b/wiki/concepts/图生视频.md @@ -0,0 +1,26 @@ +--- +title: "图生视频" +type: concept +tags: [AI视频, 生成式AI] +sources: [14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md] +last_updated: 2026-04-14 +--- + +## Definition +Image-to-Video(I2V),将静态图片转化为动态视频的AI生成技术。 + +## Core Attributes +- 输入:静态图片(jpg/png等) +- 控制方式:提示词、动作模板、运镜参数、首尾帧 +- 输出:动态视频(通常2-6秒) +- 核心能力:主体一致性、运动控制、风格化 + +## Related Concepts +- [[主体一致性]]:视频中人物/物体保持与原图一致 +- [[运镜控制]]:通过参数控制镜头运动方式 +- [[提示词控制]]:通过文本描述控制视频生成 + +## Tool Ecosystem +中国平台:[[绘蛙AI视频]]、[[智谱清影]]、[[通义万相]]、[[Vidu]]、[[可灵AI]]、[[海螺AI]]、[[即梦AI]]、[[万相营造]]、[[艺映AI]] + +海外平台:[[PixVerse]]、[[Video Ocean]]、[[Stable Video]]、[[Viva]]、[[Haiper]] diff --git a/wiki/concepts/增量备份.md b/wiki/concepts/增量备份.md new file mode 100644 index 00000000..d3ac4ce5 --- /dev/null +++ b/wiki/concepts/增量备份.md @@ -0,0 +1,46 @@ +--- +title: "增量备份" +type: concept +tags: [backup, rsync, incremental, nfs] +last_updated: 2025-11-11 +--- + +## Definition +增量备份只复制自上次备份以来变化的文件,相比全量备份节省时间和存储空间。 + +## 工具 +| 工具 | 说明 | +|------|------| +| rsync | 高效文件同步工具 | +| NFS | 网络文件系统 | +| cron | 定时任务调度 | + +## rsync关键参数 +```bash +rsync -azR --delete \ + --exclude="venv/" \ + --exclude=".venv/" \ + --exclude="**/__pycache__/" \ + --exclude=".git/" \ + /源目录/ /目标目录/ +``` + +| 参数 | 作用 | +|------|------| +| -a | 归档模式(保留权限、时间等) | +| -z | 压缩传输 | +| -R | 相对路径 | +| --delete | 删除目标中不存在的文件 | + +## 备份策略 +- 备份频率:每日凌晨3点 +- 备份目标:NAS挂载目录 +- 监控:日志写入/var/log/rsync_backup.log + +## 安全措施 +- 锁文件防止并发运行 +- 挂载点检查防止数据写错位置 +- _netdev参数确保网络就绪后再挂载 + +## 相关文档 +- [[Ubuntu服务器通过rsync实现日常增量备份]] \ No newline at end of file diff --git a/wiki/concepts/多云策略.md b/wiki/concepts/多云策略.md new file mode 100644 index 00000000..fd885292 --- /dev/null +++ b/wiki/concepts/多云策略.md @@ -0,0 +1,28 @@ +--- +title: "多云策略" +type: concept +tags: [Cloud, Multi-Cloud, Strategy] +sources: [multi-cloud-strategy-roi.md, cloud-operating-model.md] +last_updated: 2025-03-01 +--- + +## Summary +多云策略是使用多个云服务提供商以优化性能、避免供应商锁定并增强安全性的战略方法。 + +## Definition +多云策略指同时使用多个公有云提供商(如AWS、Azure、GCP)的服务,避免单一供应商依赖。 + +## Benefits +- 避免供应商锁定 +- 增加弹性和可靠性 +- 改善安全态势 +- 成本优化 +- 访问最新创新 +- 合规灵活性 +- 性能优化 + +## Connections +- [[Kubernetes]] ← 实现 ← 多云策略 +- [[Terraform]] ← 支撑 ← 多云管理 +- [[FinOps]] ← 优化 ← 多云成本 + diff --git a/wiki/concepts/异步任务处理.md b/wiki/concepts/异步任务处理.md new file mode 100644 index 00000000..02d7bb60 --- /dev/null +++ b/wiki/concepts/异步任务处理.md @@ -0,0 +1,61 @@ +--- +title: "异步任务处理" +type: concept +tags: [async, task-queue, django-q] +sources: [tiktok-pm-python-django-project.md] +last_updated: 2026-04-14 +--- + +## Definition +异步任务处理将耗时的操作(如API调用、数据导入)放到后台任务队列中执行,避免HTTP请求超时。 + +## Implementation: Django-Q + +### Configuration +```python +Q_CLUSTER = { + 'name': 'DjangORM', + 'workers': 4, + 'timeout': 360, + 'retry': 120, + 'queue_limit': 50, + 'orm': 'default', +} +``` + +### Task Function +```python +from django_q.tasks import async_task + +def trigger_bright_data_task(urls_list): + # 触发Bright Data API + job_id = trigger_bright_data_api(urls_list) + # 异步轮询结果 + async_task('products.tasks.poll_bright_data_result', job_id) + return job_id + +def poll_bright_data_result(job_id): + # 轮询直到任务完成 + while True: + status = check_job_status(job_id) + if status == 'completed': + download_and_import_data() + break + time.sleep(30) +``` + +### Worker +启动后台worker处理异步任务: +```bash +docker compose exec web python manage.py qcluster +``` + +## Bright Data Integration +- 触发异步任务获取job_id +- 每30秒轮询任务状态 +- 任务完成后下载JSON数据并导入数据库 + +## Connections +- [[异步任务处理]] ← uses ← [[Django]] +- [[异步任务处理]] ← calls ← [[Bright Data]] +- [[异步任务处理]] ← runs_in ← [[Docker容器化部署]] diff --git a/wiki/concepts/意图理解引擎.md b/wiki/concepts/意图理解引擎.md new file mode 100644 index 00000000..6c9e6364 --- /dev/null +++ b/wiki/concepts/意图理解引擎.md @@ -0,0 +1,24 @@ +--- +id: 意图理解引擎 +title: "意图理解引擎" +type: concept +tags: [ai, understanding, image-generation] +sources: [] +last_updated: 2025-12-18 +--- + +# 意图理解引擎 + +AI模型理解用户创作意图而非简单匹配关键词的能力。 + +## Definition +新一代AI图像生成模型的核心能力,能够理解创作目的、物理规则、构图美学,并做出合乎逻辑的艺术决策。 + +## Key Properties +- 理解"为什么"或"为谁"创作 +- 物理规则推演(如光影反射逻辑) +- 构图美学理解(黄金分割、视觉层次) +- 语义上下文推理(品牌调性、受众定位) + +## Related Entities +- [[Nano Banana Pro]] \ No newline at end of file diff --git a/wiki/concepts/指纹浏览器.md b/wiki/concepts/指纹浏览器.md new file mode 100644 index 00000000..bdb97e84 --- /dev/null +++ b/wiki/concepts/指纹浏览器.md @@ -0,0 +1,42 @@ +--- +title: "指纹浏览器" +type: concept +tags: [fingerprint, browser, ads-power, isolation] +last_updated: 2025-12-31 +--- + +## Definition +指纹浏览器通过模拟不同浏览器环境和设备参数,实现账号环境隔离,防止账号关联被检测。 + +## 核心原理 +- 每个浏览器环境独立Cookie/缓存 +- 模拟不同User-Agent、时区、语言 +- 配合代理IP实现环境完全隔离 + +## 关键配置 +| 配置项 | 说明 | +|------|------| +| 浏览器版本 | 选择最新Chrome | +| 操作系统 | 选择Windows | +| 代理类型 | SOCKS5 | +| 代理来源 | 系统代理设置复制 | + +## 验证步骤 +1. 检查IP是否在目标地区(ip111.cn) +2. 多网站验证IP一致性 +3. 用scamalytics.com检测IP纯净度(需低风险) + +## 适用场景 +- 多账号注册(电商、社交媒体) +- 跨境服务订阅(Claude Pro等) +- 防关联批量操作 + +## 推荐工具 +| 工具 | 说明 | +|------|------| +| AdsPower | 指纹浏览器,免费5环境 | +| PingMe | 短信接码平台 | +| WildCard | 虚拟信用卡 | + +## 相关文档 +- [[如何用指纹浏览器安全注册并订阅Claude Pro会员全攻略]] \ No newline at end of file diff --git a/wiki/concepts/提示词优化.md b/wiki/concepts/提示词优化.md new file mode 100644 index 00000000..a56c8cbb --- /dev/null +++ b/wiki/concepts/提示词优化.md @@ -0,0 +1,23 @@ +--- +id: 提示词优化 +title: "提示词优化" +type: concept +tags: [prompt-engineering, ai] +sources: [] +last_updated: 2025-12-18 +--- + +# 提示词优化 + +通过精细化设计提示词来提升AI生成内容质量的技术。 + +## Definition +在AI生成任务中,通过优化输入提示词的结构、描述精度和上下文信息,获得更符合预期的输出结果。 + +## Key Properties +- 影响生成内容质量的关键因素 +- 需要明确主体、场景、光线、氛围等细节 +- 自然语言描述优于关键词堆砌 + +## Related Concepts +- [[结构化输出]] diff --git a/wiki/concepts/提示词框架.md b/wiki/concepts/提示词框架.md new file mode 100644 index 00000000..b87f2f06 --- /dev/null +++ b/wiki/concepts/提示词框架.md @@ -0,0 +1,23 @@ +--- +title: "提示词框架" +type: concept +tags: [提示词, AI绘图, 结构化] +sources: [nano-banana-ti-shi-ci-kuang-jia.md] +last_updated: 2026-04-14 +--- + +## Definition +用于标准化AI绘图提示词结构的JSON模板,包含shot、subject、environment、lighting、camera等标准字段。 + +## Core Attributes +- 核心字段:shot、subject、environment、lighting、camera、color_grade、style、quality、negatives +- camera子字段:focal_length、aperture、angle +- negatives字段:排除不需要的元素 + +## Types +- [[物件描述框架]]:item、materials、details、condition +- [[人物描述框架]]:age、appearance、pose + +## Related Concepts +- [[Nano Banana]] ← 使用 ← [[提示词框架]] +- [[提示词优化]] ← 相关 ← [[提示词框架]] diff --git a/wiki/concepts/文字生成视频.md b/wiki/concepts/文字生成视频.md new file mode 100644 index 00000000..4ac8de98 --- /dev/null +++ b/wiki/concepts/文字生成视频.md @@ -0,0 +1,23 @@ +--- +id: 文字生成视频 +title: "文字生成视频" +type: concept +tags: [ai, video-generation, text-to-video] +sources: [] +last_updated: 2025-12-18 +--- + +# 文字生成视频 + +AI将文本描述转换为视频内容的技术。 + +## Definition +利用人工智能技术,根据用户输入的文字描述自动生成对应视频内容的技术。 + +## Related Entities +- [[万彩AI]] +- [[百度AI开放平台]] +- [[Sora]] + +## Related Concepts +- [[视频自动化生成工作流]] \ No newline at end of file diff --git a/wiki/concepts/无服务器密码管理.md b/wiki/concepts/无服务器密码管理.md new file mode 100644 index 00000000..1a60e65f --- /dev/null +++ b/wiki/concepts/无服务器密码管理.md @@ -0,0 +1,21 @@ +--- +title: "无服务器密码管理" +type: concept +tags: [cloudflare, edge-computing, password-manager] +sources: [nodewarden-ba-bitwarden-ban-dao-cloudflare-workers-shang-che-di-gao-bie-fu-wu-qi.md] +last_updated: 2026-04-14 +--- + +## Definition +无服务器密码管理是将传统密码管理服务器端逻辑部署到边缘计算平台,实现真正的零服务器运维。 + +## Key Attributes +- 平台:Cloudflare Workers +- 存储:Cloudflare D1/R2 +- 优点:无服务器费用,无VPS需求 + +## Related Entities +- [[NodeWarden]]:无服务器密码管理的实现 + +## Connections +- [[无服务器密码管理]] ← 实现 ← [[NodeWarden]] diff --git a/wiki/concepts/本地LLM.md b/wiki/concepts/本地LLM.md new file mode 100644 index 00000000..ebf7b09c --- /dev/null +++ b/wiki/concepts/本地LLM.md @@ -0,0 +1,47 @@ +--- +title: "本地LLM" +type: concept +tags: [llm, ollama, local, ai] +last_updated: 2026-04-13 +--- + +## Definition +本地LLM是在本地服务器部署运行的大语言模型,不依赖外部API服务。 + +## 优势 +- 隐私:数据不离开本地 +- 成本:无API调用费用 +- 离线:无需互联网连接 +- 定制:可微调模型 + +## Ollama部署流程 +```bash +# 1. 安装 +curl -fsSL https://ollama.com/install.sh | sh + +# 2. 下载模型 +ollama pull qwen2.5-coder:7b + +# 3. 运行 +ollama run qwen2.5-coder:7b +``` + +## 推荐模型 +| 模型 | 大小 | 用途 | +|------|------|------| +| qwen2.5-coder:7b | 4.6GB | 代码生成、DevOps | +| qwen2.5-coder:3b | ~2GB | 轻量任务 | + +## 与n8n集成 +n8n通过HTTP Request调用Ollama API: +```bash +POST http://localhost:11434/api/generate +{ + "model": "qwen2.5-coder:7b", + "prompt": "..." +} +``` + +## 相关文档 +- [[在 Ubuntu 安装 Ollama 并运行 Qwen2.5-Coder 7B]] +- [[可自动化、可扩展、AI增强的电商数据采集与处理系统]] \ No newline at end of file diff --git a/wiki/concepts/死亡过滤器.md b/wiki/concepts/死亡过滤器.md new file mode 100644 index 00000000..7d9d00a5 --- /dev/null +++ b/wiki/concepts/死亡过滤器.md @@ -0,0 +1,25 @@ +--- +title: "死亡过滤器" +type: concept +tags: [AI, 决策, 人生] +sources: [bu-tan-ji-shu] +last_updated: 2026-04-14 +--- + +## Definition +乔布斯.skill提出的人生决策工具:通过问"如果今天是最后一天,我还会做今天要做的事吗"来筛选真正重要的事。 + +## 核心问题 +"如果今天是最后一天,我还会做今天要做的事吗?" + +## 应用 +- 别问"什么AI技能最赚钱" +- 问自己对什么东西有genuine的热爱和curiosity +- 然后用AI把它做到极致 + +## 核心观点 +"对一千件事说No,只对一件事说Yes,然后把那一件事做到insanely great。" — 乔布斯.skill + +## Connections +- [[乔布斯.skill]] ← 来源 ← [[死亡过滤器]] +- [[端到端能力]] ← 行动 ← [[死亡过滤器]] diff --git a/wiki/concepts/流程工程.md b/wiki/concepts/流程工程.md new file mode 100644 index 00000000..9ca68b30 --- /dev/null +++ b/wiki/concepts/流程工程.md @@ -0,0 +1,23 @@ +--- +title: "流程工程" +type: concept +tags: [AI应用, 工作流, 方法论] +sources: [32wan-ren-shou-cang-de-claude-skills.md] +last_updated: 2026-04-14 +--- + +## Definition +从"提示词工程"升级后的AI应用新范式,强调将业务经验沉淀为SOP并交给AI稳定执行。 + +## Core Attributes +- 核心转变:从优化提示词到构建可执行流程 +- 价值:可复用、可规模化、可审计 +- 代表:Claude Skills + +## Related Concepts +- [[Claude Skills]] ← 代表案例 ← [[流程工程]] +- [[提示词工程]] ← 升级前 ← [[流程工程]] +- [[SOP]] ← 核心组成 ← [[流程工程]] + +## Key Insight +未来真正有价值的,不是谁的Prompt写得最花,而是谁最懂业务流程、谁能把经验沉淀成SOP、谁能把SOP交给AI稳定执行。 diff --git a/wiki/concepts/混合搜索.md b/wiki/concepts/混合搜索.md new file mode 100644 index 00000000..0f0dccef --- /dev/null +++ b/wiki/concepts/混合搜索.md @@ -0,0 +1,31 @@ +--- +title: "混合搜索" +type: concept +tags: [AI, 搜索, 记忆] +sources: [yang-long-xia-5-tian] +last_updated: 2026-04-14 +--- + +## Definition +结合BM25(关键词匹配)、向量嵌入(语义相似性)和重排序器的搜索方式。 + +## 搜索方式对比 + +| 类型 | 优点 | 缺点 | +|------|------|------| +| 纯BM25 | 精确匹配关键词、数字、短语 | 无法理解语义 | +| 纯向量 | 理解语义相似性 | 专有名词/数字匹配差 | +| 混合搜索 | 兼顾精确和语义 | 需要配置 | + +## QMD实现 +QMD = BM25 + 向量 + 重排序器 +- 搜索"n8n工作流执行结果"时 +- 找到包含确切词语的结果 AND 语义相关的结果 +- 按相关性重新排序 + +## 关键洞察 +"纯语义搜索理论上听起来不错,但在专有名词、具体数字和确切短语上失败。混合搜索明显更好。" — 比利哥 + +## Connections +- [[QMD]] ← 实现 ← [[混合搜索]] +- [[RAG]] ← 底层技术 ← [[混合搜索]] diff --git a/wiki/concepts/监控体系.md b/wiki/concepts/监控体系.md new file mode 100644 index 00000000..3d530f94 --- /dev/null +++ b/wiki/concepts/监控体系.md @@ -0,0 +1,39 @@ +--- +title: "监控体系" +type: concept +tags: [monitoring, prometheus, grafana, metrics] +last_updated: 2025-11-11 +--- + +## Definition +监控体系是用于收集、展示和告警系统/应用指标的完整解决方案。 + +## 核心组件 +| 组件 | 用途 | +|------|------| +| Prometheus | 时序数据库,采集和存储指标 | +| Grafana | 可视化仪表盘和告警管理 | +| Alertmanager | 告警分发(邮件/Slack/Webhook) | + +## 数据采集层(Exporters) +| Exporter | 采集内容 | 端口 | +|----------|----------|------| +| node_exporter | 主机指标(CPU/内存/磁盘/网络) | 9100 | +| cAdvisor | Docker容器指标 | 8080 | +| blackbox_exporter | HTTP/TCP/DNS探测 | 9115 | + +## 监控维度 +1. **主机层**:CPU、内存、磁盘、网络、I/O +2. **容器层**:运行状态、重启次数、资源限制 +3. **服务层**:HTTP可用性、响应码、延迟、错误率、TLS证书 +4. **日志层**:应用错误/异常(可选Loki) + +## 推荐告警规则 +- CPU使用率>85%持续2分钟 +- 磁盘剩余空间<10% +- 内存可用<15% +- HTTP探测失败连续2分钟 +- TLS证书剩余<14天 + +## 相关文档 +- [[家庭监控方案:Prometheus + Grafana + Node Exporter + cAdvisor + Blackbox]] \ No newline at end of file diff --git a/wiki/concepts/知识整理先行.md b/wiki/concepts/知识整理先行.md new file mode 100644 index 00000000..7b32fbeb --- /dev/null +++ b/wiki/concepts/知识整理先行.md @@ -0,0 +1,23 @@ +--- +title: "知识整理先行" +type: concept +tags: [AI, 简报, 知识管理] +sources: [chatgpt-knowledge-to-canva-gamma] +last_updated: 2026-04-14 +--- + +## Definition +简报制作前先进行资料研究和知识整理的流程。 + +## 四阶段流程 +1. **阶段一(5分钟)**:让AI上网搜索,调阅10笔以上资料 +2. **阶段二(1分钟)**:让AI建立知识架构,有客观认识和主观诠释 +3. **阶段三(1分钟)**:让AI输出文字版简报大纲 +4. **阶段四**:粘贴到Canva/Gamma生成版面 + +## 核心观点 +"简报不是从版面设计开始,而是从资料研究开始。" + +## Connections +- [[ChatGPT]] ← 工具 ← [[知识整理先行]] +- [[Canva]] ← 设计 ← [[知识整理先行]] diff --git a/wiki/concepts/磁盘镜像备份.md b/wiki/concepts/磁盘镜像备份.md new file mode 100644 index 00000000..04c9027d --- /dev/null +++ b/wiki/concepts/磁盘镜像备份.md @@ -0,0 +1,22 @@ +--- +title: "磁盘镜像备份" +type: concept +tags: [backup, disaster-recovery, disk] +sources: [clonezilla-dui-ubuntu-server-jin-hang-quan-pan-jing-xiang-bei-fen.md] +last_updated: 2026-04-14 +--- + +## Definition +磁盘镜像备份是将整个磁盘或分区的全部内容复制为单个镜像文件的过程,用于灾难恢复时快速还原系统状态。 + +## Key Attributes +- 备份范围:整个磁盘(savedisk)或单个分区 +- 压缩选项:-z1p(高压缩率) +- 存储格式:镜像文件(.img) + +## Related Entities +- [[Clonezilla]]:执行磁盘镜像备份的工具 +- [[Ubuntu-Server]]:常见的备份目标 + +## Connections +- [[磁盘镜像备份]] ← 工具 ← [[Clonezilla]] diff --git a/wiki/concepts/端到端能力.md b/wiki/concepts/端到端能力.md new file mode 100644 index 00000000..3a9463b0 --- /dev/null +++ b/wiki/concepts/端到端能力.md @@ -0,0 +1,26 @@ +--- +title: "端到端能力" +type: concept +tags: [AI, 职业, 思维] +sources: [bu-tan-ji-shu] +last_updated: 2026-04-14 +--- + +## Definition +从零到一构建完整产品/服务/体验的能力,而不是成为别人AI流水线上的一个零件。 + +## 核心观点 +- iPod成功不是因为有最好的MP3解码器 +- 而是从iTunes到iPod到iTunes Store的整个体验 +- **一个人用AI做出一个完整的App,比一个100人的团队里当"AI提示词工程师"强一万倍** + +## 零件思维 vs 成品思维 +| 零件思维 | 成品思维 | +|---------|---------| +| 学一个AI工具就能找到工作 | 用AI做一个完整的东西 | +| 成为别人AI流水线上的螺丝钉 | 从头到尾控制整个体验 | +| 最容易被替换 | 不可替代 | + +## Connections +- [[乔布斯.skill]] ← 来源 ← [[端到端能力]] +- [[死亡过滤器]] ← 指引 ← [[端到端能力]] diff --git a/wiki/concepts/系统提示词构建.md b/wiki/concepts/系统提示词构建.md new file mode 100644 index 00000000..4f43f3bd --- /dev/null +++ b/wiki/concepts/系统提示词构建.md @@ -0,0 +1,27 @@ +--- +id: 系统提示词构建 +title: "系统提示词构建" +type: concept +tags: [prompt-engineering, system-prompt] +sources: [] +last_updated: 2025-12-30 +--- + +# 系统提示词构建 + +定义AI助手行为准则和能力的系统化方法。 + +## Definition +通过明确的核心身份准则、沟通规范、技术规范和安全防护,构建可靠、一致的AI助手系统。 + +## Core Categories +1. **核心身份与行为准则**:项目约定、代码风格、架构模式 +2. **沟通与互动**:专业语气、Markdown格式化、不重复信息 +3. **任务执行与工作流**:TODO列表、分解任务、验证步骤 +4. **技术与编码规范**:变量命名、模块化、错误处理 +5. **安全与防护**:密钥保护、最小权限 +6. **工具使用**:并行执行、专用工具优先 + +## Related Concepts +- [[提示词优化]] +- [[结构化输出]] \ No newline at end of file diff --git a/wiki/concepts/终端代理.md b/wiki/concepts/终端代理.md new file mode 100644 index 00000000..103caa15 --- /dev/null +++ b/wiki/concepts/终端代理.md @@ -0,0 +1,23 @@ +--- +title: "终端代理" +type: concept +tags: [linux, proxy, socks5] +sources: [ubuntu-server-ke-xue-shang-wang.md] +last_updated: 2026-04-14 +--- + +## Definition +终端代理通过工具让原本不支持代理的命令通过SOCKS5通道访问网络。 + +## Key Attributes +- 工具:ProxyChains +- 配置:/etc/proxychains4.conf +- 效果:任意终端命令走代理 + +## Related Entities +- [[V2Ray/Tuic]]:提供SOCKS5代理源 +- [[ProxyChains]]:实现终端代理的工具 + +## Connections +- [[终端代理]] ← 工具 ← [[ProxyChains]] +- [[终端 \ No newline at end of file diff --git a/wiki/concepts/自然语言工作流生成.md b/wiki/concepts/自然语言工作流生成.md new file mode 100644 index 00000000..bf22b71e --- /dev/null +++ b/wiki/concepts/自然语言工作流生成.md @@ -0,0 +1,34 @@ +--- +title: "自然语言工作流生成" +type: concept +tags: [n8n, claude, AI] +sources: [n8n-Claude-通过自然语言自动化工作流.md, 使用Claude自动生成N8N工作流的实操教程.md] +last_updated: 2026-04-14 +--- + +## Definition +通过输入自然语言描述,让 AI 自动生成工作流的技术。AI 根据描述自动选择节点、编写代码、配置参数。 + +## Mechanism +1. 用户输入自然语言需求 +2. AI 理解任务并规划工作流结构 +3. AI 自动选择合适的节点 +4. AI 生成配置和代码 +5. 用户审核并微调 + +## Benefits +- 降低学习门槛,无需手动选择节点 +- 提高效率,AI 可完成 80%-90% 的布局工作 +- 适合新手快速上手 + +## Limitations +- 约有 10%-20% 的错误率需人工修正 +- 复杂逻辑仍需人工调整 + +## Related Concepts +- [[n8n-mcp]]:实现此功能的中间件 +- [[Extended Thinking]]:提升生成质量的推理模式 + +## Connections +- [[Claude]] ← 使用 ← [[自然语言工作流生成]] +- [[自然语言工作流生成]] ← 通过 ← [[n8n-mcp]] \ No newline at end of file diff --git a/wiki/concepts/视频自动化生成工作流.md b/wiki/concepts/视频自动化生成工作流.md new file mode 100644 index 00000000..a80d1474 --- /dev/null +++ b/wiki/concepts/视频自动化生成工作流.md @@ -0,0 +1,27 @@ +--- +id: 视频自动化生成工作流 +title: "视频自动化生成工作流" +type: concept +tags: [automation, video-generation, workflow] +sources: [] +last_updated: 2025-12-18 +--- + +# 视频自动化生成工作流 + +通过API接口和自动化工具实现视频生成的批量处理流程。 + +## Definition +利用AI视频生成模型(如Sora)结合工作流自动化工具(如n8n),实现视频内容的批量、自动、高效生成。 + +## Key Properties +- 降低视频制作成本 +- 提高内容生产效率 +- 支持批量生成UGC内容 + +## Related Concepts +- [[UGC创作]] +- [[提示词优化]] + +## Examples +- Sora API + n8n自动化编排 diff --git a/wiki/concepts/认知操作系统.md b/wiki/concepts/认知操作系统.md new file mode 100644 index 00000000..c697e835 --- /dev/null +++ b/wiki/concepts/认知操作系统.md @@ -0,0 +1,28 @@ +--- +title: "认知操作系统" +type: concept +tags: [AI, Skill, 思维] +sources: [yang-xia-ri-ji-5] +last_updated: 2026-04-14 +--- + +## Definition +女娲·Skill蒸馏出的每个Skill都被比喻为一个认知操作系统——用别人的脑子思考自己的人生。 + +## 核心比喻 +- 每个Skill是一个认知操作系统 +- 你不需要同意他的所有观点 +- 但你可以在需要的时候,用他的镜片看自己的问题 + +## 典型应用场景 +| Skill | 应用场景 | +|-------|---------| +| 芒格 | AI编程时 | +| 海明威 | 写作时 | +| 乔布斯 | 做产品时 | +| 苏东坡 | 逆境中保持风骨时 | +| 塔勒布 | 提升决策质量时 | + +## Connections +- [[Skill蒸馏]] ← 本质 ← [[认知操作系统]] +- [[苏东坡]] ← 实例 ← [[认知操作系统]] diff --git a/wiki/concepts/路由器科学上网.md b/wiki/concepts/路由器科学上网.md new file mode 100644 index 00000000..7f36a107 --- /dev/null +++ b/wiki/concepts/路由器科学上网.md @@ -0,0 +1,23 @@ +--- +title: "路由器科学上网" +type: concept +tags: [proxy, router, transparent-proxy] +sources: [rax50-lu-you-qi-geng-xin-merlin-clash-ding-yue.md, ubuntu-server-ke-xue-shang-wang.md] +last_updated: 2026-04-14 +--- + +## Definition +路由器科学上网是在路由器层面配置代理,实现局域网内所有设备透明代理访问。 + +## Key Attributes +- 方式:路由器插件(Merlin Clash) +- 效果:局域网设备无需单独配置代理 +- 协议:VLESS/VMess/Trojan等 + +## Related Entities +- [[Merlin Clash]]:路由器科学上网的插件实现 +- [[RAX50]]:运行科学上网的路由器 + +## Connections +- [[路由器科学上网]] ← 工具 ← [[Merlin Clash]] +- [[路由器科学上网]] ← 平台 ← [[RAX50]] diff --git a/wiki/concepts/身份锁定.md b/wiki/concepts/身份锁定.md new file mode 100644 index 00000000..f0c37769 --- /dev/null +++ b/wiki/concepts/身份锁定.md @@ -0,0 +1,23 @@ +--- +id: 身份锁定 +title: "身份锁定" +type: concept +tags: [image-generation, consistency, reference-image] +sources: [] +last_updated: 2025-12-18 +--- + +# 身份锁定 + +在图像生成中保持角色或人物面部特征一致性的技术。 + +## Definition +通过提供参考图像并明确指示模型保持面部特征一致,使得同一角色可以在不同场景中保持一致的外观。 + +## Key Properties +- Nano Banana Pro支持最多14张参考图像(6张高保真度) +- 明确说明保持与参考图像完全一致的面部特征 +- 可在保持身份的同时改变表情、动作、场景 + +## Related Entities +- [[Nano Banana Pro]] \ No newline at end of file diff --git a/wiki/concepts/远程开发.md b/wiki/concepts/远程开发.md new file mode 100644 index 00000000..457e7d7e --- /dev/null +++ b/wiki/concepts/远程开发.md @@ -0,0 +1,22 @@ +--- +title: "远程开发" +type: concept +tags: [Trae, SSH, 开发] +sources: [trae-remote-dev-guide] +last_updated: 2026-04-14 +--- + +## Definition +通过Remote-SSH连接到远程服务器进行开发的方式。 + +## SSH免密配置 +1. 生成密钥对:`ssh-keygen -t rsa -b 4096` +2. 上传公钥:`ssh-copy-id -i ~/.ssh/id_rsa.pub user@host` +3. 配置SSH Config别名方便管理 + +## Docker权限 +用户必须在docker用户组:`sudo usermod -aG docker $USER` + +## Connections +- [[Trae]] ← 工具 ← [[远程开发]] +- [[Docker]] ← 容器 ← [[远程开发]] diff --git a/wiki/concepts/递归自优化.md b/wiki/concepts/递归自优化.md new file mode 100644 index 00000000..8e07eec1 --- /dev/null +++ b/wiki/concepts/递归自优化.md @@ -0,0 +1,28 @@ +--- +title: "递归自优化" +type: concept +tags: [AI理论, 自我优化, 生成系统] +sources: [formalization-recursive-self-optimizing-generative-systems.md] +last_updated: 2026-04-14 +--- + +## Definition +通过迭代自我修改构建稳定生成能力的AI系统框架。 + +## Core Attributes +- 核心循环:生成 → 优化 → 更新 → 生成 +- 目标:收敛到稳定生成能力(固定点),而非特定最优输出 +- 组成:Generator(α)、Optimizer(Ω)、Meta-generator + +## Formalization +- Generator空间:G ⊆ P^I +- 优化算子:O: P × Ω → P +- 元生成算子:M: G × P → G +- 自映射:Φ: G → G + +## Related Concepts +- [[固定点理论]] ← 数学基础 ← [[递归自优化]] +- [[λ演算]] ← 表达形式 ← [[递归自优化]] +- [[自举]] ← 机制 ← [[递归自优化]] +- [[Generator]] ← 核心组件 ← [[递归自优化]] +- [[Optimizer]] ← 核心组件 ← [[递归自优化]] diff --git a/wiki/concepts/逻辑备份.md b/wiki/concepts/逻辑备份.md new file mode 100644 index 00000000..368278e6 --- /dev/null +++ b/wiki/concepts/逻辑备份.md @@ -0,0 +1,21 @@ +--- +title: "逻辑备份" +type: concept +tags: [backup, database, postgres] +sources: [minio-zipline-zi-cheng-guan-tu-chuang-ying-yong-an-zhuang-jiao-cheng.md] +last_updated: 2026-04-14 +--- + +## Definition +逻辑备份是通过数据库工具导出数据为SQL语句或纯文本格式,区别于直接复制数据文件的物理备份。 + +## Key Attributes +- 工具:pg_dump(PostgreSQL) +- 格式:SQL文件或压缩格式 +- 优点:跨版本兼容,迁移能力强 + +## Related Entities +- [[Zipline]]:其PostgreSQL数据库需要逻辑备份 + +## Connections +- [[逻辑备份]] ← 工具 ← [[Zipline]] diff --git a/wiki/concepts/防弹笔记法.md b/wiki/concepts/防弹笔记法.md new file mode 100644 index 00000000..804f578e --- /dev/null +++ b/wiki/concepts/防弹笔记法.md @@ -0,0 +1,24 @@ +--- +title: "防弹笔记法" +type: concept +tags: [笔记, 知识管理] +sources: [chatgpt-knowledge-to-canva-gamma] +last_updated: 2026-04-14 +--- + +## Definition +任务导向+动态演化+简单精准的笔记系统。 + +## 核心定位 +为输出而设计 + +## 最小可用模板 +- 抬头:任务名称(动词开头)|完成条件(可验收)|截止日 +- 三栏:决策记录/下一步×3/参考片段 +- 复盘区:本次做法摘要/下次改进/可复用规则 + +## SSOT +Single Source of Truth,每个任务一则笔记 + +## Connections +- [[知识整理先行]] ← 应用 ← [[防弹笔记法]] diff --git a/wiki/entities/AWS.md b/wiki/entities/AWS.md new file mode 100644 index 00000000..cbc709fa --- /dev/null +++ b/wiki/entities/AWS.md @@ -0,0 +1,24 @@ +--- +title: "AWS" +type: entity +tags: [Cloud, IaaS] +sources: [] +last_updated: 2025-03-01 +--- + +## Summary +AWS(Amazon Web Services)是亚马逊公司的云服务平台,提供IaaS、PaaS和SaaS服务。 + +## Key Claims +- AWS提供超过200种云服务 +- AWS与Azure、GCP是三大主流云提供商 +- AWS CloudFormation支持跨账户StackSets部署 + +## Aliases +- Amazon Web Services +- AWS + +## Connections +- [[多云策略]] ← 支持 ← AWS +- [[Kubernetes]] ← 托管 ← EKS + diff --git a/wiki/entities/Airtable.md b/wiki/entities/Airtable.md new file mode 100644 index 00000000..09a71956 --- /dev/null +++ b/wiki/entities/Airtable.md @@ -0,0 +1,17 @@ +--- +title: "Airtable" +type: entity +tags: [database, inventory] +sources: [n8n-full-tutorial-building-ai-agents-in-2025-for-beginners.md] +last_updated: 2026-04-14 +--- + +## Definition +Airtable 是一款在线数据库服务,支持表格视图和 API 集成。 + +## Usage +作为 n8n 的外部工具集成,Airtable 可用于库存管理和数据存储。 + +## Connections +- [[n8n]] ← 集成 ← [[Airtable]] +- [[Agentic System]] ← 使用 ← [[Airtable]] \ No newline at end of file diff --git a/wiki/entities/Alex-Ewerlof.md b/wiki/entities/Alex-Ewerlof.md new file mode 100644 index 00000000..3ac4873e --- /dev/null +++ b/wiki/entities/Alex-Ewerlof.md @@ -0,0 +1,22 @@ +--- +title: "Alex Ewerlöf" +type: entity +tags: [author, reliability-engineering, multi-agent] +last_updated: 2026-04-13 +--- + +## Profile + +- 资深工程师(Senior Staff Engineer),27年工作经验 +- 瑞典皇家理工学院(KTH)系统工程硕士学位 +- 过去十年专注于可靠性工程和弹性架构 +- 自2023年起专攻LLM领域 + +## Key Contributions + +- [[Multi-Agent System Reliability]]:提出四种架构模式提升多智能体系统可靠性 +- 可靠性工程思维(Reliability Engineering Mindset) + +## Connections + +- [[Multi-Agent System Reliability]] ← 作者 diff --git a/wiki/entities/Anthropic.md b/wiki/entities/Anthropic.md new file mode 100644 index 00000000..ff4392e9 --- /dev/null +++ b/wiki/entities/Anthropic.md @@ -0,0 +1,19 @@ +--- +title: "Anthropic" +type: entity +tags: [AI公司, Claude] +sources: [32wan-ren-shou-cang-de-claude-skills.md] +last_updated: 2026-04-14 +--- + +## Definition +AI安全公司,Claude模型的开发商,Claude Skills官方仓库发布方。 + +## Core Attributes +- 类型:AI公司 +- 产品:Claude, Claude Skills +- 官网:anthropic.com + +## Connections +- [[Claude]] ← 开发商 ← [[Anthropic]] +- [[Claude Skills]] ← 发布方 ← [[Anthropic]] diff --git a/wiki/entities/Apache-Superset.md b/wiki/entities/Apache-Superset.md new file mode 100644 index 00000000..667da7ba --- /dev/null +++ b/wiki/entities/Apache-Superset.md @@ -0,0 +1,34 @@ +--- +title: "Apache Superset" +type: entity +tags: [bi, data-visualization, python] +sources: [install-apache-superset-in-docker.md] +last_updated: 2026-04-14 +--- + +## Definition +Apache Superset是一款开源的BI(商业智能)可视化平台,支持SQL查询、图表构建和仪表板创建。 + +## Aliases +- Superset +- Apache Superset + +## Key Attributes +- 类型:BI可视化平台 +- 语言:Python +- 许可:Apache License 2.0 +- 默认端口:8088 + +## Key Capabilities +- SQL IDE,支持多数据库连接 +- 丰富的图表类型(ECharts、D3.js等) +- 拖拽式仪表板构建 +- 缓存层(Redis/Celery) + +## Related Entities +- [[Docker]]:部署方式 +- [[MySQL]]:元数据库后端 + +## Connections +- [[Apache Superset]] ← 部署于 ← [[Docker]] +- [[MySQL]] ← 元数据库 ← [[Apache Superset]] diff --git a/wiki/entities/Atlassian-Jira.md b/wiki/entities/Atlassian-Jira.md new file mode 100644 index 00000000..fa8d7892 --- /dev/null +++ b/wiki/entities/Atlassian-Jira.md @@ -0,0 +1,22 @@ +--- +title: "Atlassian Jira" +type: entity +tags: [Collaboration, Project Management] +sources: [] +last_updated: 2025-03-02 +--- + +## Summary +Atlassian Jira是一个团队协作和工作流管理平台,支持敏捷开发和DevOps实践。 + +## Key Claims +- Jira支持Scrum和Kanban敏捷框架 +- Jira提供团队间实时通信和工作流透明度 + +## Aliases +- Jira +- Atlassian Jira + +## Connections +- [[DevOps文化]] ← 工具 ← Jira + diff --git a/wiki/entities/BloopAI.md b/wiki/entities/BloopAI.md new file mode 100644 index 00000000..1cd43785 --- /dev/null +++ b/wiki/entities/BloopAI.md @@ -0,0 +1,24 @@ +--- +id: BloopAI +title: "BloopAI" +type: entity +tags: [ai, coding-agent, company] +aliases: [Bloop] +--- + +## Definition +BloopAI 是一家专注于 AI 编码工具开发的公司,开发了 [[Vibe-Kanban]] 和 [[OpenCode]] 两款开源产品。 + +## Core Properties +- **类型**:公司/组织 +- **产品**: + - [[Vibe-Kanban]]:AI 编码 Agent 任务看板 + - [[OpenCode]]:开源 AI 编码 Agent + +## Related Entities +- [[Vibe-Kanban]]:BloopAI 开发的产品 +- [[OpenCode]]:BloopAI 开发的产品 + +## Connections +- [[BloopAI]] ← 开发 ← [[Vibe-Kanban]] +- [[BloopAI]] ← 开发 ← [[OpenCode]] diff --git a/wiki/entities/Bright-Data.md b/wiki/entities/Bright-Data.md new file mode 100644 index 00000000..1059cd67 --- /dev/null +++ b/wiki/entities/Bright-Data.md @@ -0,0 +1,29 @@ +--- +title: "Bright Data" +type: entity +tags: [api, web-scraping, tiktok] +sources: [tiktok-pm-python-django-project.md] +last_updated: 2026-04-14 +--- + +## Definition +Bright Data是一个网页数据抓取平台,提供异步API来批量获取TikTok产品数据。 + +## Role in TikTok PM Project +- 通过异步请求API抓取TikTok产品JSON数据 +- 支持批量URL提交 +- 通过轮询机制检查任务完成状态 + +## API Endpoint +``` +POST https://api.brightdata.com/datasets/v3/trigger +``` + +## Key Parameters +- dataset_id:数据集标识符 +- notify:是否发送通知 +- include_errors:是否包含错误信息 + +## Connections +- [[Bright Data]] ← provides_data ← [[异步任务处理]] +- [[Bright Data]] ← outputs_to ← [[Django]] diff --git a/wiki/entities/Caddy.md b/wiki/entities/Caddy.md new file mode 100644 index 00000000..48e79198 --- /dev/null +++ b/wiki/entities/Caddy.md @@ -0,0 +1,16 @@ +--- +title: "Caddy" +type: entity +tags: [web server, reverse proxy] +sources: [n8n-docker-install-update.md] +last_updated: 2026-04-14 +--- + +## Definition +Caddy 是一款自动 HTTPS 的反向代理服务器。 + +## Usage +在 n8n Docker 部署中,Caddy 用于提供 HTTPS 访问,反向代理到 n8n 容器。 + +## Connections +- [[n8n]] ← 访问 ← [[Caddy]] ← 提供 ← [[HTTPS]] \ No newline at end of file diff --git a/wiki/entities/Canva.md b/wiki/entities/Canva.md new file mode 100644 index 00000000..9821b24f --- /dev/null +++ b/wiki/entities/Canva.md @@ -0,0 +1,21 @@ +--- +title: "Canva" +type: entity +tags: [设计, 简报, AI] +sources: [chatgpt-knowledge-to-canva-gamma] +last_updated: 2026-04-14 +--- + +## Definition +在线设计工具,支持AI简报生成,但不适合前期资料整理。 + +## Core Attributes +- 类型:设计工具/简报软件 +- AI功能:2025年推出AI问答功能,支持中文 +- 定价:免费账号提供丰富模板 + +## 正确用法 +先在ChatGPT完成知识整理和大纲,再粘贴到Canva AI生成简报版面。 + +## Connections +- [[ChatGPT]] ← 内容来源 ← [[Canva]] diff --git a/wiki/entities/ChatGPT.md b/wiki/entities/ChatGPT.md new file mode 100644 index 00000000..1ed23085 --- /dev/null +++ b/wiki/entities/ChatGPT.md @@ -0,0 +1,20 @@ +--- +title: "ChatGPT" +type: entity +tags: [AI, OpenAI, 知识管理] +sources: [chatgpt-knowledge-to-canva-gamma] +last_updated: 2026-04-14 +--- + +## Definition +OpenAI开发的AI对话模型,可用于知识整理和简报大纲生成。 + +## 在简报流程中的作用 +1. 5分钟资料研究和收集 +2. 1分钟建立知识架构 +3. 1分钟输出简报大纲 +4. 最后粘贴到Canva/Gamma制作 + +## Connections +- [[Canva]] ← 设计 ← [[ChatGPT]] +- [[Gamma]] ← 设计 ← [[ChatGPT]] diff --git a/wiki/entities/ChinaTextbook.md b/wiki/entities/ChinaTextbook.md new file mode 100644 index 00000000..33869f00 --- /dev/null +++ b/wiki/entities/ChinaTextbook.md @@ -0,0 +1,18 @@ +--- +title: "ChinaTextbook" +type: entity +tags: [资源, 教育, GitHub] +sources: [china-textbook] +last_updated: 2026-04-14 +--- + +## Definition +开源项目,收集中国小学、初中、高中、大学PDF教材。 + +## Core Attributes +- 大小:41.53 GB +- 来源:国家中小学智慧教育平台 +- 地址:github.com/TapXWorld/ChinaTextbook + +## Connections +- [[GitHub]] ← 托管 ← [[ChinaTextbook]] diff --git a/wiki/entities/Chuck-Carroll.md b/wiki/entities/Chuck-Carroll.md new file mode 100644 index 00000000..24edc607 --- /dev/null +++ b/wiki/entities/Chuck-Carroll.md @@ -0,0 +1,19 @@ +--- +title: "Chuck Carroll" +type: entity +tags: [author, tutorial, youtube, rss] +last_updated: 2025-10-10 +--- + +## Profile + +- 技术博主 +- 撰写关于YouTube RSS获取方法的教程 + +## Key Contributions + +- [[How to Get the RSS Feed For Any YouTube Channel]]:提供无需第三方服务获取YouTube频道RSS的方法 + +## Connections + +- [[How to Get the RSS Feed For Any YouTube Channel]] ← 作者 diff --git a/wiki/entities/Claude-Code-Templates.md b/wiki/entities/Claude-Code-Templates.md new file mode 100644 index 00000000..8100638c --- /dev/null +++ b/wiki/entities/Claude-Code-Templates.md @@ -0,0 +1,35 @@ +--- +id: Claude-Code-Templates +title: "Claude-Code-Templates" +type: entity +tags: [claude-code, template, skill] +aliases: [aitmpl, claude-code-templates] +--- + +## Definition +Claude-Code-Templates 是一个模板平台,提供 Claude Code 可复用的 Skills、Agents、MCP 资源集合。 + +## Core Properties +- **类型**:模板平台 +- **网址**:https://www.aitmpl.com +- **资源类型**:Skills、Agents、MCP + +## Installation +通过 npx 命令安装: +```bash +npx claude-code-templates@latest --skill=<path> --yes +``` + +## Example +安装 git-commit-helper Skill: +```bash +npx claude-code-templates@latest --skill=development/git-commit-helper --yes +``` + +## Related Entities +- [[Claude]]:Claude Code 的底层 AI 模型 + +## Connections +- [[Claude-Code-Templates]] ← 提供模板 ← [[Skill]] +- [[Claude-Code-Templates]] ← 提供模板 ← [[Agent]] +- [[Claude-Code-Templates]] ← 提供模板 ← [[MCP]] diff --git a/wiki/entities/Claude-Skills.md b/wiki/entities/Claude-Skills.md new file mode 100644 index 00000000..6e92de55 --- /dev/null +++ b/wiki/entities/Claude-Skills.md @@ -0,0 +1,19 @@ +--- +title: "Claude Skills" +type: entity +tags: [AI技能, Claude, 工作流] +sources: [32wan-ren-shou-cang-de-claude-skills.md] +last_updated: 2026-04-14 +--- + +## Definition +Anthropic官方发布的AI技能指南,本质是写给Claude的"说明书"和"SOP"。 + +## Core Attributes +- 类型:AI技能/工作流 +- 官方仓库:https://github.com/anthropics/skills(3.2万星) +- 分类:Office四大件、开发者工具箱、创意类Skill + +## Connections +- [[Claude Skills]] ← 发布方 ← [[Anthropic]] +- [[Vibe Coding]] ← 尽头 ← [[Claude Skills]] diff --git a/wiki/entities/Cline.md b/wiki/entities/Cline.md new file mode 100644 index 00000000..db9aa9ba --- /dev/null +++ b/wiki/entities/Cline.md @@ -0,0 +1,20 @@ +--- +title: "Cline" +type: entity +tags: [AI编程, 开源, VSCode插件] +sources: [2025-nai-ge-shen-ji-ai-kai-yuan-ping-ti.md] +last_updated: 2026-04-14 +--- + +## Definition +VS Code生态中最强大的开源自主编程插件,Cursor的最佳开源平替。 + +## Core Attributes +- 类型:AI编程插件 +- 平台:VS Code +- 特点:深度理解项目上下文,支持MCP扩展,敏感操作需用户授权 +- 开源地址:https://github.com/cline/cline + +## Connections +- [[AI编程]] ← 开源平替 ← [[Cline]] +- [[Cursor]] ← 原版 ← [[Cline]] diff --git a/wiki/entities/Clonezilla.md b/wiki/entities/Clonezilla.md new file mode 100644 index 00000000..d4905a61 --- /dev/null +++ b/wiki/entities/Clonezilla.md @@ -0,0 +1,33 @@ +--- +title: "Clonezilla" +type: entity +tags: [backup, cloning, linux] +sources: [clonezilla-dui-ubuntu-server-jin-hang-quan-pan-jing-xiang-bei-fen.md] +last_updated: 2026-04-14 +--- + +## Definition +Clonezilla(再生龙)是一款开源的磁盘克隆和备份工具,类似于Ghost,支持将整个磁盘或分区备份为镜像文件。 + +## Aliases +- Clonezilla Live +- 再生龙 + +## Key Attributes +- 类型:磁盘备份工具 +- 许可:开源GPL +- 支持架构:amd64, i386 +- 发行版:Debian-based + +## Key Capabilities +- 磁盘到镜像备份(savedisk) +- 磁盘到磁盘克隆(devdev) +- 支持NFS/SMB/SSH网络存储 +- 支持多种文件系统(ext4, NTFS, Btrfs等) + +## Related Entities +- [[Rufus]]:用于制作Clonezilla启动盘的工具 +- [[Ubuntu-Server]]:常见的备份目标系统 + +## Connections +- [[Clonezilla]] ← 工具 ← [[磁盘镜像备份]] diff --git a/wiki/entities/CloudDrive2.md b/wiki/entities/CloudDrive2.md new file mode 100644 index 00000000..d0a7c914 --- /dev/null +++ b/wiki/entities/CloudDrive2.md @@ -0,0 +1,19 @@ +--- +title: "CloudDrive2" +type: entity +tags: [云盘同步, NAS应用] +last_updated: 2026-04-14 +--- + +## Summary +Synology NAS 上的云盘同步应用,支持多种云存储服务。 + +## 描述 +CloudDrive2 是群晖 NAS 上的第三方应用,可将阿里云盘、115 网盘等云存储服务挂载为本地驱动器,实现 NAS 与云存储之间的同步。 + +## Aliases +- CloudDrive2 + +## 连接 +- [[Synology NAS]]:安装平台 +- [[阿里云盘]]:支持的云存储服务 diff --git a/wiki/entities/Coze.md b/wiki/entities/Coze.md new file mode 100644 index 00000000..1f71746d --- /dev/null +++ b/wiki/entities/Coze.md @@ -0,0 +1,19 @@ +--- +title: "Coze" +type: entity +tags: [AI平台, Agent, 字节跳动] +sources: [ai-jie-jue-fang-an-zhuan-jia-pei-xin-ke-cheng.md] +last_updated: 2026-04-14 +--- + +## Definition +字节跳动推出的AI Agent平台,支持国内版(coze.cn)和海外版(coze.com)。 + +## Core Attributes +- 类型:AI Agent平台 +- 厂商:字节跳动 +- 版本:国内版、海外版 +- 特点:可视化工作流、插件市场、多行业Agent模板 + +## Connections +- [[AI Agent]] ← 平台 ← [[Coze]] diff --git a/wiki/entities/Cpolar.md b/wiki/entities/Cpolar.md new file mode 100644 index 00000000..d7541385 --- /dev/null +++ b/wiki/entities/Cpolar.md @@ -0,0 +1,16 @@ +--- +title: "Cpolar" +type: entity +tags: [networking, tunnel] +sources: [n8n-configure-telegram-trigger.md] +last_updated: 2026-04-14 +--- + +## Definition +Cpolar 是一款内网穿透工具,用于将本地服务暴露到公网。 + +## Usage +使用 Cpolar 可以将本地 n8n 实例暴露到公网,使得 Telegram 等外部服务可以访问 Webhook。 + +## Connections +- [[n8n]] ← 暴露 ← [[Cpolar]] \ No newline at end of file diff --git a/wiki/entities/Cursor.md b/wiki/entities/Cursor.md new file mode 100644 index 00000000..fe3e9ca7 --- /dev/null +++ b/wiki/entities/Cursor.md @@ -0,0 +1,28 @@ +--- +title: "Cursor" +type: entity +tags: [AI, IDE, 代码编辑器] +sources: [cursor-2-0-beginner-guide, mcp-in-cursor] +last_updated: 2026-04-14 +--- + +## Definition +基于VS Code的AI增强代码编辑器,支持AI模型辅助代码生成和多任务代理操作。 + +## Core Attributes +- 类型:AI代码编辑器 +- 基础:VS Code +- 定价:免费可用,付费升级获取更多额度 +- 支持:多AI代理并行操作 + +## Key Features +- **Composer**:自研AI模型,生成速度比同类快4倍 +- **多代理模式**:同时运行不同任务,互不干扰 +- **Diff视图**:查看AI代码改动对比 +- **Agent模式**:自动执行内嵌命令并处理工具调用 +- **Ask模式**:仅提供文本答案,不改动文件 +- **MCP集成**:支持Model Context Protocol外部工具集成 + +## Connections +- [[MCP]] ← 集成 ← [[Cursor]] +- [[Composer]] ← 核心模型 ← [[Cursor]] diff --git a/wiki/entities/Dan-Koe.md b/wiki/entities/Dan-Koe.md new file mode 100644 index 00000000..53a358df --- /dev/null +++ b/wiki/entities/Dan-Koe.md @@ -0,0 +1,22 @@ +--- +title: "Dan Koe" +type: entity +tags: [author, creator-economy, personal-development] +last_updated: 2026-04-14 +--- + +## Profile + +- 内容创作者、创业者 +- 运营The Dank OE newsletter +- 倡导通才型思维和创意经济 + +## Key Contributions + +- [[If you have multiple interests, do not waste the next 2-3 years]]:关于多重兴趣转化为事业的方法论 +- 2 Hour Writer产品创始人 +- Eden软件创始人 + +## Connections + +- [[If you have multiple interests, do not waste the next 2-3 years]] ← 作者 diff --git a/wiki/entities/DeepSeek.md b/wiki/entities/DeepSeek.md new file mode 100644 index 00000000..b62ad8f5 --- /dev/null +++ b/wiki/entities/DeepSeek.md @@ -0,0 +1,17 @@ +--- +title: "DeepSeek" +type: entity +tags: [AI, 模型, AGI] +sources: [deepseek-handbook] +last_updated: 2026-04-14 +--- + +## Definition +专注于通用人工智能(AGI)的中国科技公司。 + +## Core Attributes +- 类型:AI公司 +- 产品:DeepSeek-R1开源推理模型 + +## Connections +- [[清华大学]] ← 手册发布 ← [[DeepSeek]] diff --git a/wiki/entities/Dify.md b/wiki/entities/Dify.md new file mode 100644 index 00000000..97e18b80 --- /dev/null +++ b/wiki/entities/Dify.md @@ -0,0 +1,19 @@ +--- +title: "Dify" +type: entity +tags: [LLM应用, 开源, 工作流] +sources: [2025-nai-ge-shen-ji-ai-kai-yuan-ping-ti.md] +last_updated: 2026-04-14 +--- + +## Definition +LLM应用开发平台,帮助企业和个人快速搭建带知识库的AI机器人。 + +## Core Attributes +- 类型:LLM应用开发平台 +- 特点:可视化界面,支持模型调试、提示词编排、工作流 +- 开源地址:https://github.com/langgenius/dify + +## Connections +- [[LLM应用]] ← 开发平台 ← [[Dify]] +- [[n8n]] ← 竞品 ← [[Dify]] diff --git a/wiki/entities/Django.md b/wiki/entities/Django.md new file mode 100644 index 00000000..6145f5f9 --- /dev/null +++ b/wiki/entities/Django.md @@ -0,0 +1,31 @@ +--- +title: "Django" +type: entity +tags: [python, web-framework, backend] +sources: [tiktok-pm-python-django-project.md] +last_updated: 2026-04-14 +--- + +## Definition +Django是一个高级Python Web框架,鼓励快速开发和简洁实用的设计。 + +## Role in TikTok PM Project +- 核心Web框架,处理HTTP请求和响应 +- ORM系统定义数据库模型(Product、ProductImage、ProductVideo、ProductVariation、ProductReview) +- Admin后台系统提供产品管理界面 +- REST Framework提供API接口 + +## Key Features Used +- Django ORM:Python类到数据库表的映射 +- Django Admin:自动生成管理后台 +- Django REST Framework:构建RESTful API +- Django-Q:异步任务队列 +- django-tinymce:富文本编辑器集成 +- django-filter:高级过滤功能 + +## Connections +- [[Django]] ← extends ← [[Django REST Framework]] +- [[Django]] ← uses ← [[MySQL]] +- [[Django]] ← runs_on ← [[Docker]] +- [[Django]] ← served_by ← [[Gunicorn]] +- [[Django]] ← proxied_by ← [[Nginx]] diff --git a/wiki/entities/Docker.md b/wiki/entities/Docker.md new file mode 100644 index 00000000..ad2fefc5 --- /dev/null +++ b/wiki/entities/Docker.md @@ -0,0 +1,20 @@ +--- +title: "Docker" +type: entity +tags: [容器, DevOps] +sources: [trae-remote-dev-guide, dev-experience-standards] +last_updated: 2026-04-14 +--- + +## Definition +容器平台,用于应用容器化。 + +## 在Trae开发中的角色 +- Ubuntu 2开发环境:源码挂载到容器,代码修改实时生效 +- Ubuntu 1生产环境:镜像打包,数据持久化 + +## 关键配置 +- 用户组:用户必须在docker用户组才能免sudo操作 + +## Connections +- [[Trae]] ← 开发 ← [[Docker]] diff --git a/wiki/entities/Flux.md b/wiki/entities/Flux.md new file mode 100644 index 00000000..7647152d --- /dev/null +++ b/wiki/entities/Flux.md @@ -0,0 +1,18 @@ +--- +title: "Flux" +type: entity +tags: [AI绘图, 开源, 图像生成] +sources: [2025-nai-ge-shen-ji-ai-kai-yuan-ping-ti.md] +last_updated: 2026-04-14 +--- + +## Definition +开源AI绘图模型,出自前Stable Diffusion核心团队之手,被称为"开源界的Midjourney"。 + +## Core Attributes +- 类型:AI图像生成 +- 特点:人体解剖学最正确的开源模型,能精准在图中写出指定单词 +- 开源地址:https://github.com/black-forest-labs/flux + +## Connections +- [[AI绘图]] ← 开源平替 ← [[Flux]] diff --git a/wiki/entities/Gamma.md b/wiki/entities/Gamma.md new file mode 100644 index 00000000..e8f20710 --- /dev/null +++ b/wiki/entities/Gamma.md @@ -0,0 +1,16 @@ +--- +title: "Gamma" +type: entity +tags: [AI, 简报, 设计] +sources: [chatgpt-knowledge-to-canva-gamma] +last_updated: 2026-04-14 +--- + +## Definition +AI简报工具,以专业AI简报效果著称。 + +## 正确用法 +先在ChatGPT完成知识整理和大纲,再粘贴到Gamma AI生成图文并茂简报。 + +## Connections +- [[ChatGPT]] ← 内容来源 ← [[Gamma]] diff --git a/wiki/entities/Gemini-3.md b/wiki/entities/Gemini-3.md new file mode 100644 index 00000000..b5d58e40 --- /dev/null +++ b/wiki/entities/Gemini-3.md @@ -0,0 +1,23 @@ +--- +id: Gemini-3 +title: "Gemini-3" +type: entity +tags: [google, ai-model, multimodal] +sources: [] +last_updated: 2025-12-18 +--- + +# Gemini-3 + +Google开发的多模态AI模型,支持文本、图像、视频等多种模态的理解和生成。 + +## Aliases +- Gemini 3 +- Gemini-3 + +## Definition +Google的多模态大语言模型,擅长结构化输出和应用开发。 + +## Key Properties +- 支持结构化输出 +- 可通过提示词约束模型思考 \ No newline at end of file diff --git a/wiki/entities/GitHub.md b/wiki/entities/GitHub.md new file mode 100644 index 00000000..84802f04 --- /dev/null +++ b/wiki/entities/GitHub.md @@ -0,0 +1,25 @@ +--- +title: "GitHub" +type: entity +tags: [Git, 代码托管] +sources: [git-push-connection-reset, china-textbook] +last_updated: 2026-04-14 +--- + +## Definition +全球最大代码托管平台,国内访问存在连接限制问题。 + +## Core Attributes +- 类型:代码托管平台 +- 问题:TCP连接被防火墙阻断(Connection reset) + +## Connection Issues +- `Recv failure: Connection was reset`:TCP层面中断 +- 原因:防火墙检测到流量特征,发送TCP RST包 + +## Solutions +- **代理方案**:配置Git代理走本地V2RayN +- **SSH方案**:切换到SSH协议(22端口相对稳定) + +## Connections +- [[Git]] ← 托管 ← [[GitHub]] diff --git a/wiki/entities/Gitea.md b/wiki/entities/Gitea.md new file mode 100644 index 00000000..77a330da --- /dev/null +++ b/wiki/entities/Gitea.md @@ -0,0 +1,27 @@ +--- +title: "Gitea" +type: entity +tags: [Git, 版本控制, 自托管] +sources: [yang-xia-ri-ji-3] +last_updated: 2026-04-14 +--- + +## Definition +自建Git服务,为Obsidian笔记库提供版本控制,实现"任何时候都能回溯"的变更管理。 + +## Core Attributes +- 类型:自托管Git服务 +- 用途:笔记历史版本管理、变更可溯源 +- 特性:支持Pull Request进行笔记变更Review + +## Value +- **任何时候都能回溯**:三个月前某台服务器上跑的什么服务,git log即可找到 +- **变更有据可查**:commit message记录变更来源和内容 +- **多人协作预留**:权限体系支持未来多Agent协作 + +## Connections +- [[Obsidian]] ← 版本控制 ← [[Gitea]] +- [[OpenClaw]] ← 写入接口 ← [[Gitea]] + +## Aliases +- Gitea diff --git a/wiki/entities/Google-Sheets.md b/wiki/entities/Google-Sheets.md new file mode 100644 index 00000000..601900db --- /dev/null +++ b/wiki/entities/Google-Sheets.md @@ -0,0 +1,16 @@ +--- +title: "Google Sheets" +type: entity +tags: [spreadsheet, data] +sources: [使用Claude自动生成N8N工作流的实操教程.md] +last_updated: 2026-04-14 +--- + +## Definition +Google Sheets 是 Google 的在线表格服务,支持 API 集成。 + +## Usage +作为 n8n 工作流的集成目标,用于新闻爬取后更新数据到 Google 表格。 + +## Connections +- [[n8n]] ← 写入 ← [[Google Sheets]] \ No newline at end of file diff --git a/wiki/entities/Gunicorn.md b/wiki/entities/Gunicorn.md new file mode 100644 index 00000000..87f65708 --- /dev/null +++ b/wiki/entities/Gunicorn.md @@ -0,0 +1,19 @@ +--- +title: "Gunicorn" +type: entity +tags: [wsgi-server, python, deployment] +sources: [tiktok-pm-python-django-project.md] +last_updated: 2026-04-14 +--- + +## Definition +Gunicorn(Green Unicorn)是一个Python WSGI HTTP服务器,用于在生产环境中运行Python Web应用。 + +## Role in TikTok PM Project +- 作为Django应用的WSGI服务器 +- 在Docker容器中运行 +- 监听8000端口 + +## Connections +- [[Gunicorn]] ← runs ← [[Django]] +- [[Gunicorn]] ← proxied_by ← [[Nginx]] diff --git a/wiki/entities/Haiper.md b/wiki/entities/Haiper.md new file mode 100644 index 00000000..2e2767db --- /dev/null +++ b/wiki/entities/Haiper.md @@ -0,0 +1,17 @@ +--- +title: "Haiper" +type: entity +tags: [AI视频, 图生视频, AI] +sources: [14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md] +last_updated: 2026-04-14 +--- + +## Definition +AI视频生成工具,支持免费无限次使用图生视频功能。 + +## Core Attributes +- 类型:AI视频生成 +- 特色:Discord免费无限使用,支持多种风格 + +## Connections +- [[图生视频]] ← 工具 ← [[Haiper]] diff --git a/wiki/entities/Homarr.md b/wiki/entities/Homarr.md new file mode 100644 index 00000000..66b39d43 --- /dev/null +++ b/wiki/entities/Homarr.md @@ -0,0 +1,18 @@ +--- +title: "Homarr" +type: entity +tags: [仪表盘, Docker] +sources: [yong-docker-an-zhuang-homarr.md] +last_updated: 2026-04-14 +--- + +## Definition +服务器状态仪表盘工具,Docker部署。 + +## Core Attributes +- 类型:服务器仪表盘 +- 镜像:ghcr.io/homarr-labs/homarr +- 端口:7575 + +## Connections +- [[Docker]] ← 容器 ← [[Homarr]] diff --git a/wiki/entities/HunyuanVideo.md b/wiki/entities/HunyuanVideo.md new file mode 100644 index 00000000..f8a1039e --- /dev/null +++ b/wiki/entities/HunyuanVideo.md @@ -0,0 +1,19 @@ +--- +title: "HunyuanVideo" +type: entity +tags: [AI视频, 开源, 腾讯] +sources: [2025-nai-ge-shen-ji-ai-kai-yuan-ping-ti.md] +last_updated: 2026-04-14 +--- + +## Definition +腾讯混元视频生成模型,开源界参数量最大的视频生成模型之一。 + +## Core Attributes +- 类型:AI视频生成 +- 厂商:腾讯 +- 特点:参数量大,中文Prompt理解能力强,动作连贯符合物理直觉 +- 开源地址:https://github.com/Tencent-Hunyuan/HunyuanVideo + +## Connections +- [[AI生视频]] ← 开源平替 ← [[HunyuanVideo]] diff --git a/wiki/entities/Jellyfin.md b/wiki/entities/Jellyfin.md new file mode 100644 index 00000000..6a0e3665 --- /dev/null +++ b/wiki/entities/Jellyfin.md @@ -0,0 +1,40 @@ +--- +title: "Jellyfin" +type: entity +tags: [jellyfin, media-server, nas, synology] +sources: [yong-docker-an-zhuang-jellyfin] +last_updated: 2026-04-14 +--- + +## Definition +Jellyfin是开源媒体服务器,支持电影/剧集管理、转码和播放,兼容Plex/Emby生态。 + +## 核心特性 +- 开源自托管媒体服务器 +- 支持硬件转码(Intel QuickSync、NVIDIA GPU、AMD VAAPI) +- 元数据自动刮削 +- 多平台客户端支持 + +## Docker部署配置 +```yaml +services: + jellyfin: + image: nyanmisaka/jellyfin:latest + user: "1026:100" + ports: + - 8096:8096/tcp + - 7359:7359/udp + volumes: + - /volume1/docker/jellyfin/config:/config + - /volume1/docker/jellyfin/cache:/cache + - /volume2/movie:/media + - "/volume1/TV shows:/media2" + devices: + - /dev/dri:/dev/dri + environment: + - JELLYFIN_PublishedServerUrl=http://jellyfin.ishenwei.online +``` + +## Connections +- [[Synology NAS]] ← 宿主 ← [[Jellyfin]] +- [[Docker]] ← 部署方式 ← [[Jellyfin]] diff --git a/wiki/entities/Karpathy.md b/wiki/entities/Karpathy.md new file mode 100644 index 00000000..133bb1d9 --- /dev/null +++ b/wiki/entities/Karpathy.md @@ -0,0 +1,24 @@ +--- +title: "Karpathy" +type: entity +tags: [人物, AI, 研究员] +sources: [yang-xia-ri-ji-3] +last_updated: 2026-04-14 +--- + +## Definition +AI研究员,提出了LLM Wiki概念,与RAG模式对比,强调AI增量构建持久化Wiki的价值。 + +## LLM Wiki核心观点 +- **RAG vs LLM Wiki**:RAG每次从零检索,知识不积累;LLM Wiki让AI增量构建和维护持久化Wiki +- **Obsidian Web Clipper**:浏览器插件随时采集网页文章 +- **Graph View用法**:发现孤岛页面(无页面链接指向)和知识盲区(被广泛提及但无专页) +- **Git自动同步**:AI批量改文件能力越强,越需要版本管理兜底 +- **QMD**:Wiki规模变大后的精准搜索工具 + +## Connections +- [[LLM-Wiki]] ← 提出 ← [[Karpathy]] +- [[Obsidian]] ← 工具 ← [[Karpathy]] + +## Aliases +- Karpathy diff --git a/wiki/entities/Kubernetes.md b/wiki/entities/Kubernetes.md new file mode 100644 index 00000000..fed9b093 --- /dev/null +++ b/wiki/entities/Kubernetes.md @@ -0,0 +1,27 @@ +--- +title: "Kubernetes" +type: entity +tags: [Container, Orchestration] +sources: [] +last_updated: 2025-03-01 +--- + +## Summary +Kubernetes是一个开源容器编排平台,用于自动化容器化应用的部署、扩展和管理。 + +## Key Claims +- Kubernetes支持跨云环境的工作负载编排 +- EKS、GKE、AKS分别是AWS、GCP、Azure的托管Kubernetes服务 +- Kubernetes是多云策略的关键组件 + +## Aliases +- K8s +- Kubernetes +- EKS +- GKE +- AKS + +## Connections +- [[多云策略]] ← 依赖 ← Kubernetes +- [[IaC]] ← 互补 ← 容器化 + diff --git a/wiki/entities/LanceDB.md b/wiki/entities/LanceDB.md new file mode 100644 index 00000000..20d0b5da --- /dev/null +++ b/wiki/entities/LanceDB.md @@ -0,0 +1,26 @@ +--- +title: "LanceDB" +type: entity +tags: [数据库, 向量数据库] +sources: [yang-xia-ri-ji-2] +last_updated: 2026-04-14 +--- + +## Definition +基于LanceDB的向量数据库,用于OpenClaw的长期记忆存储,支持语义搜索。 + +## Core Attributes +- 类型:向量数据库 +- 用途:存储重要的决策、用户偏好、反复使用的流程 +- 检索方式:语义搜索 + +## 在OpenClaw中的角色 +- **长期记忆层**:memory-lancedb-pro +- 与每日memory文件(短期)配合使用 + +## Connections +- [[OpenClaw]] ← 长期记忆 ← [[LanceDB]] + +## Aliases +- LanceDB +- memory-lancedb-pro diff --git a/wiki/entities/LangChain.md b/wiki/entities/LangChain.md new file mode 100644 index 00000000..ad317795 --- /dev/null +++ b/wiki/entities/LangChain.md @@ -0,0 +1,22 @@ +--- +title: "LangChain" +type: entity +tags: [RAG框架, LLM, 开发框架] +--- + +## Definition +LangChain是一个简化RAG流程的开发框架,提供Chain、Agent、Memory等抽象。 + +## Core Mechanism +- Chain:串联LLM与工具/数据源 +- Agent:自主决策执行路径 +- Memory:会话记忆管理 + +## Key Properties +- Python和JavaScript双版本 +- 丰富的内置组件 +- 活跃的开源社区 + +## Connections +- [[RAG]] ← 工具 ← [[LangChain]] +- [[LlamaIndex]] ← 类似 ← [[LangChain]] diff --git a/wiki/entities/LaunchDarkly.md b/wiki/entities/LaunchDarkly.md new file mode 100644 index 00000000..400220b9 --- /dev/null +++ b/wiki/entities/LaunchDarkly.md @@ -0,0 +1,24 @@ +--- +title: "LaunchDarkly" +type: entity +tags: [Feature Flags, DevOps, Deployment] +sources: [] +last_updated: 2025-07-26 +--- + +## Summary +LaunchDarkly是一个功能管理(Feature Management)平台,帮助团队实现渐进式发布、金丝雀部署和即时回滚。 + +## Key Claims +- LaunchDarkly使团队能够在不重新部署的情况下控制功能开关 +- 功能标志将回滚时间从数小时缩短到秒级 +- HP通过功能标志将回滚时间从数小时缩短到数分钟 +- Christian Dior将15分钟回滚缩短到即时切换 + +## Aliases +- LaunchDarkly + +## Connections +- [[RTO]] ← 改善 ← 功能标志 +- [[功能标志]] ← 平台 ← LaunchDarkly + diff --git a/wiki/entities/LlamaIndex.md b/wiki/entities/LlamaIndex.md new file mode 100644 index 00000000..fe3bdf46 --- /dev/null +++ b/wiki/entities/LlamaIndex.md @@ -0,0 +1,22 @@ +--- +title: "LlamaIndex" +type: entity +tags: [RAG框架, LLM, 开发框架] +--- + +## Definition +LlamaIndex是一个专门用于构建RAG应用的框架,专注于数据索引和检索优化。 + +## Core Mechanism +- 数据连接器:支持多种数据源 +- 索引结构:树形、列表、向量等 +- 查询引擎:支持混合检索 + +## Key Properties +- 专注RAG场景 +- 与LangChain可互操作 +- 丰富的索引策略 + +## Connections +- [[RAG]] ← 工具 ← [[LlamaIndex]] +- [[LangChain]] ← 类似 ← [[LlamaIndex]] diff --git a/wiki/entities/Mac-Mini.md b/wiki/entities/Mac-Mini.md new file mode 100644 index 00000000..ebc5e31d --- /dev/null +++ b/wiki/entities/Mac-Mini.md @@ -0,0 +1,40 @@ +--- +title: "Mac Mini" +type: entity +tags: [hardware, apple, server, arm64] +last_updated: 2026-04-13 +--- + +## Overview +Mac Mini M4是家庭网络的主控节点,运行AI助手框架和多个服务。 + +## Attributes +| 属性 | 值 | +|------|-----| +| 型号 | Mac Mini M4 | +| CPU架构 | Apple Silicon (ARM64) | +| 内网IP | 192.168.3.189 | +| 公共域名 | macmini.ishenwei.online | +| SSH | 已启用 | + +## 运行的服务 +| 服务 | 端口 | 说明 | +|------|------|------| +| OpenClaw | 8080 | AI助手框架 | +| vaultwarden | 5151 | 密码管理 | +| stq_nginx | 7777 | STQ项目反向代理 | +| stq_frontend | 5173 | STQ前端 | +| stq_web | 8000 | STQ Web服务 | +| stq_mariadb | 3306 | STQ数据库 | +| stq-n8n | 62000 | STQ n8n | + +## FRP端口映射 +| 名称 | 类型 | localPort | remotePort | +|------|------|------------|------------| +| macmini-ssh | tcp | 22 | 60026 | +| vaultwarden | tcp | 5151 | 15151 | + +## 相关文档 +- [[Mac Mini 安装 FRP 0.65.0(ARM64)操作笔记]] +- [[Mac Mini 服务器配置:防止自动锁屏与睡眠]] +- [[家庭网络环境概览 2026-04-03]] \ No newline at end of file diff --git a/wiki/entities/Manus.md b/wiki/entities/Manus.md new file mode 100644 index 00000000..b96feb26 --- /dev/null +++ b/wiki/entities/Manus.md @@ -0,0 +1,19 @@ +--- +title: "Manus" +type: entity +tags: [AI智能体, AI公司] +sources: [2025-nai-ge-shen-ji-ai-kai-yuan-ping-ti.md] +last_updated: 2026-04-14 +--- + +## Definition +AI Agent领域的年度现象级产品,2025年定义AI Agent元年的里程碑式存在,已被Meta以数十亿美金收购。 + +## Core Attributes +- 类型:AI智能体产品 +- 收购:Meta,数十亿美金 +- 意义:定义AI Agent元年 + +## Connections +- [[AI智能体]] ← 里程碑产品 ← [[Manus]] +- [[OpenManus]] ← 开源平替 ← [[Manus]] diff --git a/wiki/entities/Merlin-Clash.md b/wiki/entities/Merlin-Clash.md new file mode 100644 index 00000000..a1056dd4 --- /dev/null +++ b/wiki/entities/Merlin-Clash.md @@ -0,0 +1,32 @@ +--- +title: "Merlin Clash" +type: entity +tags: [clash, merlin-firmware, router] +sources: [rax50-lu-you-qi-geng-xin-merlin-clash-ding-yue.md] +last_updated: 2026-04-14 +--- + +## Definition +Merlin Clash是基于Clash的梅林固件科学上网插件,支持在ASUS路由器上运行。 + +## Aliases +- Merlin Clash +- 梅林Clash + +## Key Attributes +- 类型:路由器科学上网插件 +- 基础:Clash +- 固件:Asuswrt-Merlin + +## Key Capabilities +- 支持VLESS/VMess/Trojan等协议 +- 订阅自动更新 +- 局域网透明代理 +- 规则分流 + +## Related Entities +- [[RAX50]]:支持的路由器型号 +- [[V2Ray/Tuic]]:提供代理节点 + +## Connections +- [[Merlin Clash]] ← 运行于 ← [[RAX50]] diff --git a/wiki/entities/MinIO.md b/wiki/entities/MinIO.md new file mode 100644 index 00000000..87437a4a --- /dev/null +++ b/wiki/entities/MinIO.md @@ -0,0 +1,35 @@ +--- +title: "MinIO" +type: entity +tags: [s3, object-storage, self-hosted] +sources: [minio-zipline-zi-cheng-guan-tu-chuang-ying-yong-an-zhuang-jiao-cheng.md] +last_updated: 2026-04-14 +--- + +## Definition +MinIO是一款高性能的S3兼容对象存储服务器,可自托管部署。 + +## Aliases +- MinIO Server +- MinIO + +## Key Attributes +- 类型:S3兼容对象存储 +- 语言:Go +- 许可:AGPLv3 +- API端口:9000 +- Console端口:9001 + +## Key Capabilities +- S3 API完全兼容 +- 高性能(可达到每秒GB级吞吐) +- 分布式存储扩展 +- Kubernetes原生支持 +- 兼容所有S3客户端 + +## Related Entities +- [[Zipline]]:使用MinIO作为存储后端 +- [[Synology-NAS]]:常见部署平台 + +## Connections +- [[MinIO]] ← 存储后端 ← [[Zipline]] diff --git a/wiki/entities/MiniMax-M2.7.md b/wiki/entities/MiniMax-M2.7.md new file mode 100644 index 00000000..d4b63017 --- /dev/null +++ b/wiki/entities/MiniMax-M2.7.md @@ -0,0 +1,28 @@ +--- +title: "MiniMax-M2.7" +type: entity +tags: [AI, 模型, 200K-Context] +sources: [yang-xia-ri-ji-4, yang-long-xia-5-tian] +last_updated: 2026-04-14 +--- + +## Definition +MiniMax推出的AI模型,拥有200K上下文窗口,是OpenClaw的主力模型。 + +## Core Attributes +- 类型:AI模型 +- Context Window:200K tokens +- 用途:OpenClaw默认主力模型 + +## 与deepseek-reasoner对比 +| 模型 | Context Window | 用途 | +|------|---------------|------| +| MiniMax-M2.7 | 200K | 主力模型 | +| deepseek-reasoner | 16K | 备用/调试中出现问题 | + +## Connections +- [[OpenClaw]] ← 主力模型 ← [[MiniMax-M2.7]] +- [[deepseek-reasoner]] ← 对比 ← [[MiniMax-M2.7]] + +## Aliases +- MiniMax-M2.7 diff --git a/wiki/entities/MySQL.md b/wiki/entities/MySQL.md new file mode 100644 index 00000000..facfa477 --- /dev/null +++ b/wiki/entities/MySQL.md @@ -0,0 +1,24 @@ +--- +title: "MySQL" +type: entity +tags: [database, relational-database] +sources: [tiktok-pm-python-django-project.md] +last_updated: 2026-04-14 +--- + +## Definition +MySQL是一个开源关系型数据库管理系统,广泛用于Web应用的数据存储。 + +## Role in TikTok PM Project +- 存储TikTok产品的核心数据 +- 支持utf8mb4字符集以处理emoji和复杂字符 +- 通过Django ORM进行数据操作 + +## Key Configuration +- 数据库名:tiktok_product_db +- 端口:3306 +- 字符集:utf8mb4 +- sql_mode:STRICT_TRANS_TABLES + +## Connections +- [[MySQL]] ← stores ← [[Django]] diff --git a/wiki/entities/Nano-Banana-Pro.md b/wiki/entities/Nano-Banana-Pro.md new file mode 100644 index 00000000..fdd5b030 --- /dev/null +++ b/wiki/entities/Nano-Banana-Pro.md @@ -0,0 +1,32 @@ +--- +id: Nano-Banana-Pro +title: "Nano Banana Pro" +type: entity +tags: [google, image-generation, ai-model] +sources: [] +last_updated: 2025-12-18 +--- + +# Nano Banana Pro + +Google生成式AI团队的图像生成模型,专注于专业级内容生产。 + +## Aliases +- Nano Banana Pro +- Nano-Banana Pro + +## Definition +Google的图像生成模型,在文本渲染、角色一致性、视觉合成、世界知识和4K高分辨率输出方面表现出色。 + +## Key Properties +- 意图理解引擎突破 +- 物理规则推演能力 +- 构图美学理解 +- 支持最多14张参考图像 +- 利用Google搜索实现信息锚定 +- 支持1K至4K原生分辨率 + +## Related Concepts +- [[身份锁定]] +- [[意图理解引擎]] +- [[信息锚定]] \ No newline at end of file diff --git a/wiki/entities/Navidrome.md b/wiki/entities/Navidrome.md new file mode 100644 index 00000000..66e88ed9 --- /dev/null +++ b/wiki/entities/Navidrome.md @@ -0,0 +1,18 @@ +--- +title: "Navidrome" +type: entity +tags: [音乐, 流媒体, Docker] +sources: [yong-docker-zhong-an-zhuang-navidrome.md] +last_updated: 2026-04-14 +--- + +## Definition +开源音乐流媒体服务器,通过Docker部署。 + +## Core Attributes +- 类型:音乐流媒体服务器 +- 镜像:deluan/navidrome:latest +- 端口:4533 + +## Connections +- [[Docker]] ← 容器 ← [[Navidrome]] diff --git a/wiki/entities/Nginx.md b/wiki/entities/Nginx.md new file mode 100644 index 00000000..f6323d9a --- /dev/null +++ b/wiki/entities/Nginx.md @@ -0,0 +1,36 @@ +--- +title: "Nginx" +type: entity +tags: [web-server, reverse-proxy] +sources: [tiktok-pm-python-django-project.md] +last_updated: 2026-04-14 +--- + +## Definition +Nginx是一个高性能的HTTP服务器和反向代理。 + +## Role in TikTok PM Project +- 作为反向代理,将请求转发给Gunicorn +- 提供静态文件服务(/static/和/media/) +- 监听80端口 + +## Configuration +```nginx +upstream web { + server web:8000; +} + +server { + listen 80; + location /static/ { + alias /app/staticfiles/; + } + location / { + proxy_pass http://web; + } +} +``` + +## Connections +- [[Nginx]] ← proxies_to ← [[Gunicorn]] +- [[Nginx]] ← serves_static ← [[Django]] diff --git a/wiki/entities/Node-js.md b/wiki/entities/Node-js.md new file mode 100644 index 00000000..83b6152d --- /dev/null +++ b/wiki/entities/Node-js.md @@ -0,0 +1,32 @@ +--- +id: Node-js +title: "Node.js" +type: entity +tags: [javascript, runtime, environment] +aliases: [Node] +--- + +## Definition +Node.js 是一个基于 Chrome V8 引擎的 JavaScript 运行时环境,允许在服务器端执行 JavaScript 代码。 + +## Core Properties +- **类型**:运行时环境 +- **官方网址**:https://nodejs.org +- **版本**:Vibe-Kanban 推荐使用 Node 20 LTS + +## Installation Methods +- **直接安装**:从官网下载安装包 +- **nvm**:通过 Node Version Manager 管理多版本 +- **包管理器**:`sudo apt install nodejs npm`(不推荐,可能版本过旧) + +## Related Entities +- [[nvm]]:Node.js 版本管理器 +- [[PM2]]:基于 Node.js 的进程管理器 +- [[Vibe-Kanban]]:依赖 Node.js 运行 +- [[OpenCode]]:依赖 Node.js 运行 + +## Connections +- [[Node-js]] ← 安装工具 ← [[nvm]] +- [[Vibe-Kanban]] ← 依赖 ← [[Node-js]] +- [[OpenCode]] ← 依赖 ← [[Node-js]] +- [[PM2]] ← 依赖 ← [[Node-js]] diff --git a/wiki/entities/NodeWarden.md b/wiki/entities/NodeWarden.md new file mode 100644 index 00000000..3492260e --- /dev/null +++ b/wiki/entities/NodeWarden.md @@ -0,0 +1,40 @@ +--- +title: "NodeWarden" +type: entity +tags: [bitwarden, cloudflare, password-manager, serverless] +sources: [nodewarden-ba-bitwarden-ban-dao-cloudflare-workers-shang-che-di-gao-bie-fu-wu-qi.md] +last_updated: 2026-04-14 +--- + +## Definition +NodeWarden是将Bitwarden服务器端运行在Cloudflare Workers上的开源实现,实现真正无服务器的密码管理。 + +## Aliases +- NodeWarden + +## Key Attributes +- 类型:Bitwarden服务器替代品 +- 平台:Cloudflare Workers +- 存储:Cloudflare D1 + R2 +- 许可:开源 + +## Key Capabilities +- 单用户密码保管库(登录/笔记/卡片/身份) +- 全量/api/sync兼容 +- TOTP二次验证(无需付费会员) +- 附件上传/下载(基于R2) +- passkey支持 + +## Key Differences from Bitwarden +- 单用户(无组织/集合/多成员) +- 无SSO/SCIM/企业目录 +- 无Send/紧急访问功能 +- 纯免费,无订阅 + +## Related Entities +- [[Bitwarden]]:兼容的客户端 +- [[Cloudflare]]:部署平台 + +## Connections +- [[NodeWarden]] ← 服务端 ← [[Bitwarden]] +- [[Cloudflare]] ← 平台 ← [[NodeWarden]] diff --git a/wiki/entities/Node_js.md b/wiki/entities/Node_js.md new file mode 100644 index 00000000..922e3dd8 --- /dev/null +++ b/wiki/entities/Node_js.md @@ -0,0 +1,16 @@ +--- +title: "Node.js" +type: entity +tags: [runtime, javascript] +sources: [n8n-Claude-通过自然语言自动化工作流.md, 使用Claude自动生成N8N工作流的实操教程.md] +last_updated: 2026-04-14 +--- + +## Definition +Node.js 是一个基于 Chrome V8 引擎的 JavaScript 运行时,用于构建服务器端和网络应用。 + +## Usage +n8n-mcp 需要 Node.js 环境运行,npx n8n-mcp 命令依赖 Node.js。 + +## Connections +- [[n8n-mcp]] ← 运行在 ← [[Node.js]] \ No newline at end of file diff --git a/wiki/entities/NotebookLM.md b/wiki/entities/NotebookLM.md new file mode 100644 index 00000000..ed721b70 --- /dev/null +++ b/wiki/entities/NotebookLM.md @@ -0,0 +1,20 @@ +--- +title: "NotebookLM" +type: entity +tags: [AI工具, Google, 笔记, 学习] +sources: [7-ways-notebooklm-make-life-easier.md, notebooklm-open-source-alternatives.md] +last_updated: 2026-04-14 +--- + +## Definition +Google推出的AI笔记助手,通过source-grounding机制确保答案准确。 + +## Core Attributes +- 类型:AI笔记与研究助手 +- 厂商:Google +- 核心机制:source-grounding(知识库严格限制在用户上传文档内) +- 特色功能:Audio Overviews(双人播客)、项目规划、文档分析 + +## Connections +- [[Source-grounding]] ← 核心机制 ← [[NotebookLM]] +- [[Audio Overviews]] ← 功能 ← [[NotebookLM]] diff --git a/wiki/entities/Obsidian.md b/wiki/entities/Obsidian.md new file mode 100644 index 00000000..eedb8e09 --- /dev/null +++ b/wiki/entities/Obsidian.md @@ -0,0 +1,39 @@ +--- +title: "Obsidian" +type: entity +tags: [笔记工具, 知识管理] +sources: [yang-xia-ri-ji-3, yang-long-xia-5-tian] +last_updated: 2026-04-14 +--- + +## Definition +本地笔记管理工具,通过双链和Graph View支持知识网络化,配合OpenClaw实现AI输出的持久化。 + +## Core Attributes +- 类型:笔记管理工具 +- 同步方式:iCloud Drive(三端同步:Mac mini、Laptop、iPhone) +- 插件:Obsidian Git(自动commit+push)、Web Clipper(网页剪藏) + +## Key Features +- Graph View:图谱视图发现孤岛页面和知识盲区 +- 双链语法:`[[PageName]]`建立页面间连接 +- Callout块:`>[!NOTE]`标记待确认/已确认信息 +- 附件本地化:图片存储在attachments/目录 + +## Directory Structure(in nexus workspace) +``` +/Users/weishen/Workspace/nexus/ +├── openclaw/ +│ ├── knowledgebase/ # 跨Agent共用知识 +│ ├── xingshu/ # 星枢专属笔记 +│ ├── xinghui/ # 星辉专属笔记 +│ └── xingyao/ # 星曜专属笔记 +└── …(其他Obsidian笔记) +``` + +## Connections +- [[OpenClaw]] → 写入接口 → [[Obsidian]] +- [[Gitea]] → 版本控制 → [[Obsidian]] + +## Aliases +- Obsidian diff --git a/wiki/entities/Ollama.md b/wiki/entities/Ollama.md new file mode 100644 index 00000000..ac009fcb --- /dev/null +++ b/wiki/entities/Ollama.md @@ -0,0 +1,39 @@ +--- +title: "Ollama" +type: entity +tags: [ollama, local-llm, ai, python] +last_updated: 2026-04-13 +--- + +## Overview +Ollama是本地大语言模型运行平台,支持一键部署和管理LLM。 + +## Attributes +| 属性 | 值 | +|------|-----| +| 安装方式 | 官方安装脚本 | +| 默认端口 | 11434 | +| 默认监听 | 127.0.0.1 | +| 模型存储 | ~/.ollama/models | + +## 支持的模型 +| 模型 | 大小 | 用途 | +|------|------|------| +| qwen2.5-coder:7b | 4.6GB | 代码生成 | +| qwen2.5-coder:3b | ~2GB | 轻量代码生成 | + +## API接口 +- REST API: http://localhost:11434 +- API列表: http://localhost:11434/api/tags +- 聊天: http://localhost:11434/api/chat + +## 远程访问配置 +编辑`/etc/systemd/system/ollama.service`: +```ini +Environment="OLLAMA_HOST=0.0.0.0" +``` +然后重启服务使其他机器可访问。 + +## 相关文档 +- [[在 Ubuntu 安装 Ollama 并运行 Qwen2.5-Coder 7B]] +- [[可自动化、可扩展、AI增强的电商数据采集与处理系统]] \ No newline at end of file diff --git a/wiki/entities/OpenClaw.md b/wiki/entities/OpenClaw.md new file mode 100644 index 00000000..c0fae71a --- /dev/null +++ b/wiki/entities/OpenClaw.md @@ -0,0 +1,39 @@ +--- +title: "OpenClaw" +type: entity +tags: [AI, Agent, 系统软件] +sources: [yang-xia-ri-ji-1, yang-xia-ri-ji-2, yang-xia-ri-ji-3, yang-xia-ri-ji-4, yang-xia-ri-ji-5, yang-long-xia-5-tian] +last_updated: 2026-04-14 +--- + +## Definition +AI Agent操作系统,运行在Mac Mini等设备上,通过Telegram等渠道与用户交互,管理多个专业化Agent协同工作。 + +## Core Attributes +- 类型:AI Agent操作系统 +- 运行平台:Mac Mini(中央控制节点) +- 通信渠道:Telegram +- Agent数量:4个(星枢、星辉、星曜等) + +## Capabilities +- obsidian skill:笔记的读取、搜索、创建、修改 +- cron任务:定时执行自动任务 +- 记忆系统:短期记忆文件 + LanceDB向量数据库 +- self-improving:结构化经验记录与复盘 +- 多渠道路由:支持Telegram等不同渠道的模型配置 + +## Key Files +- `openclaw.json`:全局配置文件 +- `AGENTS.md`:Agent定义和启动序列 +- `memory/YYYY-MM-DD.md`:每日对话记录 +- `learnings/LEARNINGS.md`:错误规则记录 + +## Connections +- [[Obsidian]] ← 笔记载体 ← [[OpenClaw]] +- [[Gitea]] ← 版本控制 ← [[OpenClaw]] +- [[比利哥]] ← 创建者 ← [[OpenClaw]] +- [[星辉]]/[[星枢]] ← 实例 ← [[OpenClaw]] + +## Aliases +- OpenClaw +- openclaw diff --git a/wiki/entities/OpenCode.md b/wiki/entities/OpenCode.md new file mode 100644 index 00000000..8d5adc38 --- /dev/null +++ b/wiki/entities/OpenCode.md @@ -0,0 +1,43 @@ +--- +id: OpenCode +title: "OpenCode" +type: entity +tags: [ai-agent, coding-agent, open-source] +aliases: [opencode-ai] +--- + +## Definition +OpenCode 是由 [[BloopAI]] 开发的开源 AI 编码 Agent,提供终端界面、桌面应用和 IDE 扩展三种形态。支持任意 LLM Provider,可通过 API Key 配置。 + +## Core Properties +- **类型**:产品/项目 +- **开发商**:[[BloopAI]] +- **开源**:是 +- **官网**:https://opencode.ai +- **安装命令**:`curl -fsSL https://opencode.ai/install | bash` + +## Key Features +- 支持任意 LLM Provider(通过 API Key 配置) +- 提供 OpenCode Zen 托管服务(预配置模型) +- Plan 模式和 Build 模式切换(Tab 键) +- 支持 /undo 和 /redo 命令 +- 支持拖拽图片到终端 +- 支持对话分享 +- 主题定制、快捷键定制、代码格式化器定制 +- 自定义命令功能 + +## Usage Patterns +1. 初始化项目:`opencode` 进入项目后运行 `/init`,生成 `AGENTS.md` 文件 +2. 创建计划:使用 Tab 键切换到 Plan 模式 +3. 构建代码:使用 Tab 键切换到 Build 模式 +4. 代码问答:使用 `@` 键模糊搜索项目文件 + +## Related Entities +- [[BloopAI]]:开发商 +- [[Vibe-Kanban]]:集成了 OpenCode 作为支持的编码 Agent +- [[AGENTS-md]]:OpenCode 初始化时生成的项目说明文件 + +## Connections +- [[OpenCode]] ← 开发 ← [[BloopAI]] +- [[Vibe-Kanban]] ← 支持 ← [[OpenCode]] +- [[OpenCode]] ← 生成 ← [[AGENTS-md]] diff --git a/wiki/entities/OpenManus.md b/wiki/entities/OpenManus.md new file mode 100644 index 00000000..c4ca1329 --- /dev/null +++ b/wiki/entities/OpenManus.md @@ -0,0 +1,21 @@ +--- +title: "OpenManus" +type: entity +tags: [AI智能体, 开源, 工作流] +sources: [2025-nai-ge-shen-ji-ai-kai-yuan-ping-ti.md] +last_updated: 2026-04-14 +--- + +## Definition +Manus的开源平替项目,5万Star,核心逻辑是规划→执行→循环反馈。 + +## Core Attributes +- 类型:AI智能体 +- Star数:5万 +- 核心逻辑:Planning → Execution → 循环反馈 +- 技术:browser-use/Playwright,Python沙盒环境 +- 开源地址:https://github.com/FoundationAgents/OpenManus + +## Connections +- [[AI智能体]] ← 开源平替 ← [[OpenManus]] +- [[Manus]] ← 原版 ← [[OpenManus]] diff --git a/wiki/entities/OpenSSH.md b/wiki/entities/OpenSSH.md new file mode 100644 index 00000000..b479ce7b --- /dev/null +++ b/wiki/entities/OpenSSH.md @@ -0,0 +1,33 @@ +--- +title: "OpenSSH" +type: entity +tags: [remote-access, ssh, security] +sources: [ubuntu-24-04-enable-ssh.md] +last_updated: 2026-04-14 +--- + +## Definition +OpenSSH是SSH协议的开源实现,提供安全的远程登录和文件传输功能。 + +## Aliases +- OpenSSH Server +- sshd + +## Key Attributes +- 类型:远程访问工具 +- 协议:SSH +- 默认端口:22 +- 许可:BSD + +## Key Capabilities +- 加密的远程登录 +- 密钥认证支持 +- SFTP/SCP文件传输 +- 端口转发/隧道 +- Ubuntu 24.04支持ssh.socket和ssh.service两种模式 + +## Related Entities +- [[Ubuntu-Server]]:常见的运行系统 + +## Connections +- [[OpenSSH]] ← 安装于 ← [[Ubuntu-Server]] diff --git a/wiki/entities/PM2.md b/wiki/entities/PM2.md new file mode 100644 index 00000000..3ea81a08 --- /dev/null +++ b/wiki/entities/PM2.md @@ -0,0 +1,42 @@ +--- +--- +id: PM2 +title: "PM2" +type: entity +tags: [node-js, process-manager, devops] +aliases: [pm2-process-manager] +--- + +## Definition +PM2 是 Node.js 应用程序的生产级进程管理器,提供进程守护、负载均衡、自动重启和开机自启功能。 + +## Core Properties +- **类型**:工具/进程管理器 +- **技术栈**:Node.js +- **安装命令**:`npm install -g pm2` +- **官网**:http://pm2.io + +## Key Features +- 进程守护(daemonize) +- 自动重启 +- 负载均衡(cluster mode) +- 开机自启(`pm2 startup` + `pm2 save`) +- 进程监控(`pm2 monit`) +- 日志管理(`pm2 logs`) +- 进程列表(`pm2 list`) + +## Usage with Vibe-Kanban +```bash +pm2 start "HOST=0.0.0.0 PORT=9999 npx vibe-kanban" --name vibe-kanban +pm2 logs vibe-kanban +pm2 restart vibe-kanban +pm2 delete vibe-kanban +``` + +## Related Entities +- [[Vibe-Kanban]]:PM2 管理的应用之一 +- [[Node-js]]:PM2 管理的运行时环境 + +## Connections +- [[PM2]] ← 管理进程 ← [[Vibe-Kanban]] +- [[PM2]] ← 依赖 ← [[Node-js]] diff --git a/wiki/entities/Perplexica.md b/wiki/entities/Perplexica.md new file mode 100644 index 00000000..a4a2b225 --- /dev/null +++ b/wiki/entities/Perplexica.md @@ -0,0 +1,20 @@ +--- +title: "Perplexica" +type: entity +tags: [AI搜索, 开源, 本地化] +sources: [2025-nai-ge-shen-ji-ai-kai-yuan-ping-ti.md] +last_updated: 2026-04-14 +--- + +## Definition +Perplexity的开源平替,完全本地化的AI搜索引擎,2.8K Star。 + +## Core Attributes +- 类型:AI搜索引擎 +- 特点:完全本地化,支持SearXNG和本地LLM,保护隐私 +- Star数:2.8K +- 开源地址:https://github.com/ItzCrazyKns/Perplexica + +## Connections +- [[AI搜索]] ← 开源平替 ← [[Perplexica]] +- [[Perplexity]] ← 原版 ← [[Perplexica]] diff --git a/wiki/entities/PixVerse.md b/wiki/entities/PixVerse.md new file mode 100644 index 00000000..4cb89c21 --- /dev/null +++ b/wiki/entities/PixVerse.md @@ -0,0 +1,18 @@ +--- +title: "PixVerse" +type: entity +tags: [AI视频, 图生视频, AI] +sources: [14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md] +last_updated: 2026-04-14 +--- + +## Definition +爱诗科技开发的AI视频生成工具,支持多种风格和首尾帧生成。 + +## Core Attributes +- 类型:AI视频生成 +- 厂商:爱诗科技 +- 特色:多种风格(真实/动漫/3D),角色一致性 + +## Connections +- [[图生视频]] ← 工具 ← [[PixVerse]] diff --git a/wiki/entities/Plex.md b/wiki/entities/Plex.md new file mode 100644 index 00000000..a5b2ca39 --- /dev/null +++ b/wiki/entities/Plex.md @@ -0,0 +1,30 @@ +--- +title: "Plex" +type: entity +tags: [plex, media, server, streaming] +last_updated: 2025-02-23 +--- + +## Overview +Plex Media Server是家庭媒体服务器,支持多平台客户端播放视频、音乐等内容。 + +## Attributes +| 属性 | 值 | +|------|-----| +| 类型 | 媒体服务器软件 | +| 部署位置 | Synology NAS | +| 账号 | Apple ID: ishenwei@hotmail.com | + +## 客户端设备 +| 设备 | 系统 | APK位置 | +|------|------|----------| +| 当贝盒子 | Android | NAS/Software/家庭影视平台/Plex | +| 华为盒子 | Android 5.0+ | 最高支持版本10.5.0 | + +## 媒体库配置 +- 目录:aliyun-movie, aliyun-tvshows, aliyun-documentory +- 通过Xiaoya选择资源,转存到相应目录 +- Plex自动刮削并展示 + +## 相关文档 +- [[Synology NAS + Xiaoya Alist + CloudDrive2 + Plex 家庭影视平台]] \ No newline at end of file diff --git a/wiki/entities/Portainer.md b/wiki/entities/Portainer.md new file mode 100644 index 00000000..ba5e0387 --- /dev/null +++ b/wiki/entities/Portainer.md @@ -0,0 +1,33 @@ +--- +title: "Portainer" +type: entity +tags: [portainer, docker, container-management] +sources: [yong-docker-an-zhuang-portainer] +last_updated: 2026-04-14 +--- + +## Definition +Portainer是Docker可视化容器管理平台,提供Web界面管理容器、卷、网络和镜像。 + +## 核心特性 +- Docker容器可视化管理的Web UI +- 支持容器、镜像、卷、网络管理 +- 支持Edge Agent用于分布式集群管理 +- 提供CE和BE版本 + +## Docker Compose部署 +```yaml +services: + portainer: + container_name: portainer + image: portainer/portainer-ce:lts + volumes: + - /var/run/docker.sock:/var/run/docker.sock + - portainer_data:/data + ports: + - 9443:9443 + - 8000:8000 +``` + +## Connections +- [[Docker]] ← 管理对象 ← [[Portainer]] diff --git a/wiki/entities/QMD.md b/wiki/entities/QMD.md new file mode 100644 index 00000000..386093ec --- /dev/null +++ b/wiki/entities/QMD.md @@ -0,0 +1,29 @@ +--- +title: "QMD" +type: entity +tags: [搜索工具, Markdown] +sources: [yang-xia-ri-ji-3, yang-long-xia-5-tian] +last_updated: 2026-04-14 +--- + +## Definition +完全本地运行的Markdown搜索引擎,结合BM25(关键词)、向量嵌入和重排序器。 + +## Core Attributes +- 类型:本地Markdown搜索引擎 +- 地址:github.com/tobi/qmd +- 搜索方式:混合搜索(BM25 + 向量 + 重排序) + +## 与OpenClaw的结合 +解决内置SQLite搜索在精确匹配上表现差的问题: +- 纯语义搜索在专有名词、具体数字、确切短语上失败 +- 混合搜索结合关键词匹配和语义相似性 + +## 判断标准 +Wiki到几百个页面之前,index.md目录文件足够用;等AI找东西开始变慢,再接入QMD。 + +## Connections +- [[OpenClaw]] ← 搜索后端 ← [[QMD]] + +## Aliases +- QMD diff --git a/wiki/entities/Qdrant.md b/wiki/entities/Qdrant.md new file mode 100644 index 00000000..3e026df4 --- /dev/null +++ b/wiki/entities/Qdrant.md @@ -0,0 +1,23 @@ +--- +title: "Qdrant" +type: entity +tags: [向量数据库, RAG, 搜索] +--- + +## Definition +Qdrant是一个开源的向量数据库(Vector Store),用于存储Embedding Vector并实现相似度检索。 + +## Core Mechanism +- 高效存储高维向量 +- 支持相似度搜索(ANN算法) +- 提供过滤和分桶能力 + +## Key Properties +- 开源 +- 支持密集向量和稀疏向量 +- 适合RAG场景 + +## Connections +- [[Vector Store]] ← 实例 ← [[Qdrant]] +- [[Embedding Vector]] ← 存储 ← [[Qdrant]] +- [[RAG]] ← 支持 ← [[Qdrant]] diff --git a/wiki/entities/RAG.md b/wiki/entities/RAG.md new file mode 100644 index 00000000..f0a2e027 --- /dev/null +++ b/wiki/entities/RAG.md @@ -0,0 +1,27 @@ +--- +title: "RAG" +type: entity +tags: [RAG, LLM, 检索增强生成] +--- + +## Definition +RAG = Retrieval Augmented Generation,将LLM与外部数据源连接的通用方法。 + +## Core Mechanism +通过三阶段流程:Indexing → Retrieval → Generation,使LLM能够访问私有或最新数据。 + +## Key Properties +- 突破LLM Context Window限制 +- 支持私有/最新数据访问 +- 可与预训练知识互补 + +## Aliases +- Retrieval Augmented Generation +- 检索增强生成 + +## Connections +- [[Indexing]] ← 包含于 ← [[RAG]] +- [[Retrieval]] ← 包含于 ← [[RAG]] +- [[Generation]] ← 包含于 ← [[RAG]] +- [[LLM]] ← 扩展 ← [[RAG]] +- [[Vector Store]] ← 依赖 ← [[RAG]] diff --git a/wiki/entities/RAX50.md b/wiki/entities/RAX50.md new file mode 100644 index 00000000..1f0d906b --- /dev/null +++ b/wiki/entities/RAX50.md @@ -0,0 +1,32 @@ +--- +title: "RAX50" +type: entity +tags: [netgear, router, wifi-6] +sources: [rax50-lu-you-qi-geng-xin-merlin-clash-ding-yue.md] +last_updated: 2026-04-14 +--- + +## Definition +Netgear RAX50是一款支持WiFi 6的家用路由器,可刷梅林固件扩展功能。 + +## Aliases +- Netgear RAX50 +- RAX50 +- Nighthawk RAX50 + +## Key Attributes +- 类型:WiFi 6路由器 +- 厂商:Netgear +- 固件:支持Asuswrt-Merlin + +## Key Capabilities +- WiFi 6(802.11ax)支持 +- 可刷梅林固件 +- 支持Merlin Clash科学上网插件 +- USB接口用于存储共享 + +## Related Entities +- [[Merlin Clash]]:在RAX50上运行的插件 + +## Connections +- [[RAX50]] ← 运行 ← [[Merlin Clash]] diff --git a/wiki/entities/Rufus.md b/wiki/entities/Rufus.md new file mode 100644 index 00000000..33bc198e --- /dev/null +++ b/wiki/entities/Rufus.md @@ -0,0 +1,30 @@ +--- +title: "Rufus" +type: entity +tags: [bootable-usb, tool, windows] +sources: [clonezilla-dui-ubuntu-server-jin-hang-quan-pan-jing-xiang-bei-fen.md] +last_updated: 2026-04-14 +--- + +## Definition +Rufus是一款Windows平台的开源启动盘制作工具,可将ISO镜像写入U盘创建可启动介质。 + +## Aliases +- Rufus USB Creator + +## Key Attributes +- 类型:启动盘制作工具 +- 平台:Windows +- 许可:开源GPL + +## Key Capabilities +- 支持ISO写入USB设备 +- 支持GPT/MBR分区方案选择 +- 支持UEFI/BIOS启动模式 +- 可制作Clonezilla、Ubuntu等Linux发行版启动盘 + +## Related Entities +- [[Clonezilla]]:常见的 Rufus 写入目标 + +## Connections +- [[Rufus]] ← 用于制作 ← [[Clonezilla]] diff --git a/wiki/entities/RustDesk.md b/wiki/entities/RustDesk.md new file mode 100644 index 00000000..7c93faeb --- /dev/null +++ b/wiki/entities/RustDesk.md @@ -0,0 +1,19 @@ +--- +title: "RustDesk" +type: entity +tags: [远程桌面, 开源软件] +last_updated: 2026-04-14 +--- + +## Summary +开源远程桌面软件,支持自托管。 + +## 描述 +RustDesk 是一款开源的远程桌面软件,类似于 TeamViewer,支持 Windows、Linux、macOS 等平台。它允许用户通过 ID 和密码直接连接,或者部署自己的中继服务器来实现更安全的远程访问。 + +## Aliases +- RustDesk + +## 连接 +- [[Ubuntu]]:常用的远程控制目标系统 +- [[X11]]:依赖的显示协议 diff --git a/wiki/entities/Slack.md b/wiki/entities/Slack.md new file mode 100644 index 00000000..222c79b4 --- /dev/null +++ b/wiki/entities/Slack.md @@ -0,0 +1,21 @@ +--- +title: "Slack" +type: entity +tags: [Collaboration, Communication] +sources: [] +last_updated: 2025-03-02 +--- + +## Summary +Slack是一个团队实时通信和协作平台,支持DevOps团队的透明沟通。 + +## Key Claims +- Slack提供团队间实时通信和工作流可见性 +- Slack支持与多种DevOps工具集成 + +## Aliases +- Slack + +## Connections +- [[DevOps文化]] ← 工具 ← Slack + diff --git a/wiki/entities/Sora.md b/wiki/entities/Sora.md new file mode 100644 index 00000000..7ff72e4c --- /dev/null +++ b/wiki/entities/Sora.md @@ -0,0 +1,22 @@ +--- +id: Sora +title: "Sora" +type: entity +tags: [openai, video-generation, ai-model] +sources: [] +last_updated: 2025-12-18 +--- + +# Sora + +OpenAI开发的视频生成模型,能够根据文本描述生成高质量视频内容。 + +## Aliases +- Sora + +## Definition +AI视频生成模型,支持文本到视频、图像到视频等多种生成模式。 + +## Key Properties +- 成本比OpenAI其他方案便宜六倍以上 +- 支持API调用 \ No newline at end of file diff --git a/wiki/entities/Stable-Diffusion.md b/wiki/entities/Stable-Diffusion.md new file mode 100644 index 00000000..b0c6b167 --- /dev/null +++ b/wiki/entities/Stable-Diffusion.md @@ -0,0 +1,19 @@ +--- +title: "Stable Diffusion" +type: entity +tags: [AI绘图, 开源] +sources: [2025-nai-ge-shen-ji-ai-kai-yuan-ping-ti.md] +last_updated: 2026-04-14 +--- + +## Definition +老牌开源AI绘图模型,LoRA和ControlNet生态最丰富。 + +## Core Attributes +- 类型:AI图像生成 +- 特点:LoRA和ControlNet生态最丰富,适合动漫角色、特定姿势控制 +- 开源地址:https://github.com/CompVis/stable-diffusion + +## Connections +- [[AI绘图]] ← 开源项目 ← [[Stable Diffusion]] +- [[Flux]] ← 竞品 ← [[Stable Diffusion]] diff --git a/wiki/entities/Stable-Video.md b/wiki/entities/Stable-Video.md new file mode 100644 index 00000000..eff3a070 --- /dev/null +++ b/wiki/entities/Stable-Video.md @@ -0,0 +1,18 @@ +--- +title: "Stable Video" +type: entity +tags: [AI视频, 图生视频, AI] +sources: [14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md] +last_updated: 2026-04-14 +--- + +## Definition +Stability AI推出的AI视频生成平台,支持精细摄像机控制。 + +## Core Attributes +- 类型:AI视频生成 +- 厂商:Stability AI +- 特色:LoRA摄像机控制,多种风格,帧插值技术 + +## Connections +- [[图生视频]] ← 工具 ← [[Stable Video]] diff --git a/wiki/entities/Synology-NAS.md b/wiki/entities/Synology-NAS.md new file mode 100644 index 00000000..257ac978 --- /dev/null +++ b/wiki/entities/Synology-NAS.md @@ -0,0 +1,37 @@ +--- +title: "Synology NAS" +type: entity +tags: [nas, synology, storage, docker] +last_updated: 2026-04-03 +--- + +## Overview +Synology NAS DS718是家庭媒体中心和数据存储中枢,运行多个Docker服务。 + +## Attributes +| 属性 | 值 | +|------|-----| +| 型号 | DS718 | +| 内网IP | 192.168.3.17 | +| 公共域名 | nas.ishenwei.online | +| SSH | 已启用 | +| DSM版本 | 7+ | + +## 运行的服务 +| 服务 | 端口 | 说明 | +|------|------|------| +| DSM | 5000 | 管理系统 | +| Calibre | 8083 | 电子书库 | +| MinIO | 9001 | 对象存储 | +| Zipline | 3333 | 文件分享/图床 | +| Navidrome | 4533 | 音乐流媒体 | +| Jellyfin | 8096 | 媒体服务器 | +| Prometheus | 9090 | 监控系统 | +| Alertmanager | 9093 | 告警中心 | +| v2raya | 2017 | V2Ray代理 | +| vaultwarden | 5151 | 密码管理 | +| CloudDrive2 | 19798 | 云盘挂载 | + +## 相关文档 +- [[Synology NAS + Xiaoya Alist + CloudDrive2 + Plex 家庭影视平台]] +- [[家庭网络环境概览 2026-04-03]] \ No newline at end of file diff --git a/wiki/entities/Telegram.md b/wiki/entities/Telegram.md new file mode 100644 index 00000000..15e6f160 --- /dev/null +++ b/wiki/entities/Telegram.md @@ -0,0 +1,19 @@ +--- +title: "Telegram" +type: entity +tags: [messaging, webhook] +sources: [n8n-configure-telegram-trigger.md] +last_updated: 2026-04-14 +--- + +## Definition +Telegram 是一款跨平台的即时通讯服务,支持 Bot API 和 Webhook 集成。 + +## Usage +通过 Telegram Bot API,n8n 等自动化工具可以接收和处理 Telegram 消息。 + +## Related Concepts +- [[Webhook HTTPS]]:Telegram 要求 Webhook URL 必须为 HTTPS + +## Connections +- [[n8n]] ← 接收 ← [[Telegram]] \ No newline at end of file diff --git a/wiki/entities/Templater.md b/wiki/entities/Templater.md new file mode 100644 index 00000000..b4d7d0b2 --- /dev/null +++ b/wiki/entities/Templater.md @@ -0,0 +1,23 @@ +--- +title: "Templater" +type: entity +tags: [Obsidian, 插件, 模板] +sources: [templater-plugin] +last_updated: 2026-04-14 +--- + +## Definition +Obsidian的模板插件,支持变量插入和JavaScript执行。 + +## Core Attributes +- 类型:Obsidian插件 +- 开发者:SilentVoid13 +- 许可:GNU AGPLv3 + +## Key Features +- 模板语言:插入变量和函数结果 +- JavaScript支持:执行任意JS代码操作变量 +- 警告:执行任意代码可能有安全风险 + +## Connections +- [[Obsidian]] ← 插件 ← [[Templater]] diff --git a/wiki/entities/Terraform.md b/wiki/entities/Terraform.md new file mode 100644 index 00000000..6951494a --- /dev/null +++ b/wiki/entities/Terraform.md @@ -0,0 +1,24 @@ +--- +title: "Terraform" +type: entity +tags: [IaC, Cloud] +sources: [] +last_updated: 2025-03-01 +--- + +## Summary +Terraform是HashiCorp开发的开源基础设施即代码(IaC)工具,支持多云资源编排。 + +## Key Claims +- Terraform使用HCL语言定义云资源 +- Terraform支持AWS、Azure、GCP等多云提供商 +- Terraform是云无关的IaC解决方案 + +## Aliases +- Terraform +- HashiCorp Terraform + +## Connections +- [[IaC]] ← 工具 ← Terraform +- [[多云策略]] ← 支撑 ← Terraform + diff --git a/wiki/entities/TinyMCE.md b/wiki/entities/TinyMCE.md new file mode 100644 index 00000000..391f4a90 --- /dev/null +++ b/wiki/entities/TinyMCE.md @@ -0,0 +1,31 @@ +--- +title: "TinyMCE" +type: entity +tags: [rich-text-editor, javascript, django] +sources: [tiktok-pm-python-django-project.md] +last_updated: 2026-04-14 +--- + +## Definition +TinyMCE是一个基于JavaScript的富文本编辑器,可集成到Web应用中提供类似Word的编辑体验。 + +## Role in TikTok PM Project +- 集成到Django Admin中 +- 用于编辑product的description、description_1、description_2字段 +- 提供富文本格式(粗体、斜体、列表等) + +## Installation +```bash +pip install django-tinymce +``` + +## Usage +在Django Admin的ProductAdminForm中指定字段使用TinyMCE widget: +```python +widgets = { + 'description': TinyMCE(attrs={'cols': 80, 'rows': 20}), +} +``` + +## Connections +- [[TinyMCE]] ← integrated_into ← [[Django Admin定制]] diff --git a/wiki/entities/Trae.md b/wiki/entities/Trae.md new file mode 100644 index 00000000..794e4bec --- /dev/null +++ b/wiki/entities/Trae.md @@ -0,0 +1,28 @@ +--- +title: "Trae" +type: entity +tags: [IDE, Remote-SSH, 开发工具] +sources: [trae-remote-dev-guide] +last_updated: 2026-04-14 +--- + +## Definition +支持VS Code插件生态的IDE,通过Remote-SSH支持远程服务器开发。 + +## Core Attributes +- 类型:IDE +- 特点:原生支持VS Code插件生态 +- 连接方式:Remote-SSH + +## Architecture +- Ubuntu 2 (Dev Server):存放源码,运行Docker容器(代码挂载) +- Ubuntu 1 (Prod Server):运行Docker容器(镜像打包),数据持久化 +- ThinkBook (Local):仅作为UI端 + +## Development Modes +- **模式A**:Attach到正在运行的容器(推荐调试) +- **模式B**:远程编辑宿主机文件 + Docker CLI(适合编排) + +## Connections +- [[Docker]] ← 容器 ← [[Trae]] +- [[Ubuntu]] ← 服务器 ← [[Trae]] diff --git a/wiki/entities/Transmission.md b/wiki/entities/Transmission.md new file mode 100644 index 00000000..abab7e3c --- /dev/null +++ b/wiki/entities/Transmission.md @@ -0,0 +1,38 @@ +--- +title: "Transmission" +type: entity +tags: [transmission, bt, download, docker] +sources: [yong-docker-an-zhuang-transmission] +last_updated: 2026-04-14 +--- + +## Definition +Transmission是开源BT下载客户端,提供远程Web界面管理BT下载任务。 + +## 核心特性 +- 开源BT下载客户端 +- Web UI远程管理 +- 支持认证(USER/PASS环境变量) +- 支持Peer端口配置(TCP/UDP) + +## Docker部署配置 +```yaml +services: + transmission: + image: lscr.io/linuxserver/transmission:latest + ports: + - "9091:9091" # Web UI + - "51413:51413" # Peer TCP + - "51413:51413/udp" # Peer UDP + environment: + - PUID=1000 + - PGID=1000 + - USER=shenwei + - PASS=zmkm99zmkm00 + volumes: + - /home/shenwei/Docker/transmission/data:/config + - /home/shenwei/Downloads:/downloads +``` + +## Connections +- [[Docker]] ← 部署方式 ← [[Transmission]] diff --git a/wiki/entities/Ubuntu-Server.md b/wiki/entities/Ubuntu-Server.md new file mode 100644 index 00000000..b63db4bb --- /dev/null +++ b/wiki/entities/Ubuntu-Server.md @@ -0,0 +1,54 @@ +--- +title: "Ubuntu Server" +type: entity +tags: [ubuntu, server, linux, docker] +last_updated: 2026-04-03 +--- + +## Overview +家庭网络中运行的两台Ubuntu Server服务器,分别承担应用服务和开发任务。 + +## Ubuntu Server 1 +| 属性 | 值 | +|------|-----| +| 内网IP | 192.168.3.47 | +| 域名 | ubuntu1.ishenwei.online | +| SSH | 60022 (FRP映射) | + +**服务**:Grafana, Prometheus, n8n, Superset, Homarr, Transmission, Portainer, it-tools + +## Ubuntu Server 2 +| 属性 | 值 | +|------|-----| +| 内网IP | 192.168.3.45 | +| 域名 | ubuntu2.ishenwei.online | +| SSH | 60024 (FRP映射) | + +**服务**:n8n, drawio, Gitea, Portainer, it-tools, tiktok_pm + +## FRP端口映射 (Ubuntu1) +| 名称 | remotePort | +|------|------------| +| ubuntu1-ssh | 60022 | +| transmission | 19091 | +| grafana | 13000 | +| homarr | 17575 | +| superset | 18777 | +| tk | 18888 | +| ubuntu1-portainer | 19443 | +| it-tools | 18999 | + +## FRP端口映射 (Ubuntu2) +| 名称 | remotePort | +|------|------------| +| ubuntu2-ssh | 60024 | +| tk-dev | 18889 | +| n8n | 15679 | +| it-tools | 18999 | +| drawio | 18085 | + +## 相关文档 +- [[在 Ubuntu 安装 Ollama 并运行 Qwen2.5-Coder 7B]] +- [[Ubuntu服务器通过rsync实现日常增量备份]] +- [[家庭监控方案:Prometheus + Grafana + Node Exporter + cAdvisor + Blackbox]] +- [[通过VPS+内网反向代理实现域名访问内网穿透]] \ No newline at end of file diff --git a/wiki/entities/V2Ray-Tuic.md b/wiki/entities/V2Ray-Tuic.md new file mode 100644 index 00000000..6454c3c9 --- /dev/null +++ b/wiki/entities/V2Ray-Tuic.md @@ -0,0 +1,17 @@ +--- +title: "V2Ray/Tuic" +type: entity +tags: [proxy, networking] +sources: [n8n-docker-install-update.md] +last_updated: 2026-04-14 +--- + +## Definition +V2Ray 和 Tuic 是网络代理工具,支持 SOCKS5 代理协议。 + +## Usage +在 n8n Docker 部署中,V2Ray/Tuic 作为宿主机代理,通过 SOCKS5 让容器内流量科学上网。 + +## Connections +- [[Docker]] ← 网络通过 ← [[V2Ray/Tuic]] +- [[n8n]] ← 代理通过 ← [[V2Ray/Tuic]] \ No newline at end of file diff --git a/wiki/entities/VPS.md b/wiki/entities/VPS.md new file mode 100644 index 00000000..d65ab8fc --- /dev/null +++ b/wiki/entities/VPS.md @@ -0,0 +1,34 @@ +--- +title: "VPS" +type: entity +tags: [vps, server, reverse-proxy, frp] +last_updated: 2026-04-13 +--- + +## Overview +VPS(Virtual Private Server)作为公网中转服务器,运行FRPS和Caddy提供内网穿透和反向代理服务。 + +## Attributes +| 属性 | 值 | +|------|-----| +| 提供商 | RackNerd | +| 公网IP | 192.227.222.142 | +| 域名 | vps.ishenwei.online | +| SSH | ssh vps1 | +| 操作系统 | Ubuntu | + +## 运行的服务 +| 服务 | 说明 | 端口 | +|------|------|------| +| Caddy | HTTPS反向代理 | 80/443 | +| FRP Server | 内网穿透服务端 | 7000 | +| FRP Dashboard | FRP管理面板 | 7500 | + +## 架构位置 +``` +Internet → VPS (192.227.222.142) → 内网FRP客户端 +``` + +## 相关文档 +- [[通过VPS+内网反向代理实现域名访问内网穿透]] +- [[家庭网络环境概览 2026-04-03]] \ No newline at end of file diff --git a/wiki/entities/Vibe-Kanban.md b/wiki/entities/Vibe-Kanban.md new file mode 100644 index 00000000..c121327d --- /dev/null +++ b/wiki/entities/Vibe-Kanban.md @@ -0,0 +1,42 @@ +--- +id: Vibe-Kanban +title: "Vibe-Kanban" +type: entity +tags: [ai-agent, coding-agent, task-board] +aliases: [] +--- + +## Definition +Vibe-Kanban 是由 [[BloopAI]] 开发的开源 AI 编码 Agent 任务看板工具。通过看板界面管理 AI Agent 的任务执行,每个任务运行在独立的 git worktree 中隔离进行。 + +## Core Properties +- **类型**:产品/项目 +- **开发商**:[[BloopAI]] +- **技术栈**:Node.js +- **开源**:是(GitHub:https://github.com/BloopAI/vibe-kanban) +- **官网**:https://www.vibekanban.com + +## Key Features +- AI 编码 Agent 任务看板 +- 隔离的 git worktree 执行环境 +- 默认使用 `--dangerously-skip-permissions` 标志运行 +- 支持 [[MCP]] 集成 +- 支持 [[GitHub]] CLI 集成(创建 Pull Request) +- 可通过 `PORT` 环境变量指定端口 +- 可通过 `HOST` 环境变量指定绑定地址(`0.0.0.0` 允许远程访问) + +## Dependencies +- [[Node-js]](推荐 LTS 版本) +- [[OpenCode]](支持的编码 Agent 之一) +- [[PM2]](可选,用于进程管理) + +## Related Entities +- [[OpenCode]]:Vibe-Kanban 支持的编码 Agent +- [[PM2]]:可选的进程管理器 +- [[BloopAI]]:开发商 + +## Connections +- [[Vibe-Kanban]] ← 支持 ← [[OpenCode]] +- [[Vibe-Kanban]] ← 安装依赖 ← [[Node-js]] +- [[Vibe-Kanban]] ← 进程管理 ← [[PM2]] +- [[Vibe-Kanban]] ← 开发 ← [[BloopAI]] diff --git a/wiki/entities/Video-Ocean.md b/wiki/entities/Video-Ocean.md new file mode 100644 index 00000000..9c827556 --- /dev/null +++ b/wiki/entities/Video-Ocean.md @@ -0,0 +1,18 @@ +--- +title: "Video Ocean" +type: entity +tags: [AI视频, 图生视频, AI] +sources: [14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md] +last_updated: 2026-04-14 +--- + +## Definition +潞晨科技推出的多功能AI视频生成平台,支持多种画风切换。 + +## Core Attributes +- 类型:AI视频生成 +- 厂商:潞晨科技 +- 特色:V2.0版本高清逼真,多种画风 + +## Connections +- [[图生视频]] ← 工具 ← [[Video Ocean]] diff --git a/wiki/entities/Vidu.md b/wiki/entities/Vidu.md new file mode 100644 index 00000000..dfc870bc --- /dev/null +++ b/wiki/entities/Vidu.md @@ -0,0 +1,18 @@ +--- +title: "Vidu" +type: entity +tags: [AI视频, 图生视频, AI] +sources: [14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md] +last_updated: 2026-04-14 +--- + +## Definition +生数科技联合清华大学发布的AI视频大模型,中国首个长时长、高一致性、高动态性视频模型。 + +## Core Attributes +- 类型:AI视频生成 +- 机构:生数科技+清华大学 +- 特色:多主体参考功能,10秒生成一段视频 + +## Connections +- [[图生视频]] ← 工具 ← [[Vidu]] diff --git a/wiki/entities/Viva.md b/wiki/entities/Viva.md new file mode 100644 index 00000000..92c4bc9a --- /dev/null +++ b/wiki/entities/Viva.md @@ -0,0 +1,18 @@ +--- +title: "Viva" +type: entity +tags: [AI视频, 图生视频, AI] +sources: [14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md] +last_updated: 2026-04-14 +--- + +## Definition +智象未来推出的免费AI创意视觉生成平台,图生视频质量在免费产品中表现优异。 + +## Core Attributes +- 类型:AI创意视觉生成 +- 厂商:智象未来 +- 特色:完前免费,支持6种运镜方式 + +## Connections +- [[图生视频]] ← 工具 ← [[Viva]] diff --git a/wiki/entities/YouTube.md b/wiki/entities/YouTube.md new file mode 100644 index 00000000..75b70a36 --- /dev/null +++ b/wiki/entities/YouTube.md @@ -0,0 +1,17 @@ +--- +title: "YouTube" +type: entity +tags: [视频平台, n8n] +sources: [youtube-channel-id] +last_updated: 2026-04-14 +--- + +## Definition +全球最大视频平台,可通过Channel ID接入n8n等自动化工作流。 + +## Channel ID +- 格式:`UCPlwvN0w4qFSP1FllALB92w` +- 获取方法:view-source页面查询channel_id参数 + +## Connections +- [[n8n]] ← 集成 ← [[YouTube]] diff --git a/wiki/entities/Yuri-Pessa.md b/wiki/entities/Yuri-Pessa.md new file mode 100644 index 00000000..344b7eb0 --- /dev/null +++ b/wiki/entities/Yuri-Pessa.md @@ -0,0 +1,19 @@ +--- +title: "Yuri Pessa" +type: entity +tags: [author, product-design, agentic-ai] +last_updated: 2025-03-02 +--- + +## Profile + +- LinkedIn AI产品设计领域作者 +- 专注于Agentic AI产品设计研究 + +## Key Contributions + +- [[Designing for Agentic AI]]:提出Agentic AI设计的五大最佳实践原则 + +## Connections + +- [[Designing for Agentic AI]] ← 作者 diff --git a/wiki/entities/Zipline.md b/wiki/entities/Zipline.md new file mode 100644 index 00000000..ec21feba --- /dev/null +++ b/wiki/entities/Zipline.md @@ -0,0 +1,37 @@ +--- +title: "Zipline" +type: entity +tags: [image-hosting, self-hosted, s3] +sources: [minio-zipline-zi-cheng-guan-tu-chuang-ying-yong-an-zhuang-jiao-cheng.md] +last_updated: 2026-04-14 +--- + +## Definition +Zipline是一款自托管图床应用,提供图片上传API,支持与n8n等工具集成。 + +## Aliases +- Zipline Image Host +- Zipline + +## Key Attributes +- 类型:图片托管服务 +- 平台:Node.js +- 数据库:PostgreSQL/SQLite +- 存储后端:S3兼容存储 +- 默认端口:3000 + +## Key Capabilities +- 图片上传API +- API Token认证 +- 可配置上传规则 +- 支持与n8n工作流集成 +- 提供仪表板管理 + +## Related Entities +- [[MinIO]]:S3兼容存储后端 +- [[n8n]]:可调用Zipline API的工作流工具 +- [[Synology-NAS]]:部署平台 + +## Connections +- [[Zipline]] ← 存储 ← [[MinIO]] +- [[Zipline]] ← 集成 ← [[n8n]] diff --git a/wiki/entities/b_top_plus_plus.md b/wiki/entities/b_top_plus_plus.md new file mode 100644 index 00000000..42ddb6e9 --- /dev/null +++ b/wiki/entities/b_top_plus_plus.md @@ -0,0 +1,23 @@ +--- +title: "btop++" +type: entity +tags: [Linux, 系统监控, TUI] +sources: [linux-system-monitor] +last_updated: 2026-04-14 +--- + +## Definition +最好的TUI系统资源监控应用,界面美观,功能平衡。 + +## Core Attributes +- 类型:TUI系统监控 +- 安装:Pacman或Snap包 +- 特点:响应快、SSH友好 + +## 启动 +``` +btop +``` + +## Connections +- [[Linux]] ← 监控 ← [[btop++]] diff --git a/wiki/entities/deepseek-reasoner.md b/wiki/entities/deepseek-reasoner.md new file mode 100644 index 00000000..b637bd1b --- /dev/null +++ b/wiki/entities/deepseek-reasoner.md @@ -0,0 +1,31 @@ +--- +title: "deepseek-reasoner" +type: entity +tags: [AI, 模型, 16K-Context] +sources: [yang-xia-ri-ji-4] +last_updated: 2026-04-14 +--- + +## Definition +DeepSeek推出的推理模型,Context Window只有16K,是导致星枢Telegram频道Context Limit问题的根源。 + +## Core Attributes +- 类型:AI模型 +- Context Window:16K tokens +- 问题:safeguard模式预留一半token给compaction,实际可用为0 + +## 问题原因 +星枢的Telegram channel配置了deepseek-reasoner,加上OpenClaw的safeguard模式预留16384 tokens,导致: +``` +estimatedPromptTokens=393 +overflowTokens=392 +reserveTokens=16384 +# 实际可用 = 16K - 16K = 0 +``` + +## Connections +- [[星枢]] ← 绑定导致问题 ← [[deepseek-reasoner]] +- [[MiniMax-M2.7]] ← 应切换回 ← [[deepseek-reasoner]] + +## Aliases +- deepseek-reasoner diff --git a/wiki/entities/it-tools.md b/wiki/entities/it-tools.md new file mode 100644 index 00000000..1b1c07c1 --- /dev/null +++ b/wiki/entities/it-tools.md @@ -0,0 +1,35 @@ +--- +title: "it-tools" +type: entity +tags: [it-tools, developer-tools, docker] +sources: [yong-docker-an-zhuang-it-tools] +last_updated: 2026-04-14 +--- + +## Definition +it-tools是面向IT从业者的常用工具Web集合,提供URL编码/解码、JSON格式化、密码生成等在线工具。 + +## 核心特性 +- 面向IT从业者的工具集合 +- Docker容器化部署,内存限制128M +- 支持交互模式(stdin_open + tty) +- 开源免费 + +## Docker部署配置 +```yaml +services: + it-tools: + image: corentinth/it-tools:latest + container_name: it-tools + stdin_open: true + tty: true + ports: + - "8999:80" + deploy: + resources: + limits: + memory: 128M +``` + +## Connections +- [[Docker]] ← 部署方式 ← [[it-tools]] diff --git a/wiki/entities/n8n.md b/wiki/entities/n8n.md new file mode 100644 index 00000000..194834b1 --- /dev/null +++ b/wiki/entities/n8n.md @@ -0,0 +1,36 @@ +--- +title: "n8n" +type: entity +tags: [工作流, 自动化, AI] +sources: [n8n-configure-telegram-trigger.md, n8n-docker-install-update.md, n8n-full-tutorial-building-ai-agents-in-2025-for-beginners.md, n8n-Claude-通过自然语言自动化工作流.md, 使用Claude自动生成N8N工作流的实操教程.md] +last_updated: 2026-04-14 +--- + +## Definition +n8n 是一款开源的工作流自动化工具,支持通过节点连接执行任务,可通过 Docker 部署或本地安装。 + +## Properties +- **部署方式**:Docker Compose、本地安装 +- **节点数量**:543 个节点(包括 AI 节点 271 个) +- **集成能力**:支持 Telegram、Airtable、Google Sheets 等外部服务 +- **AI 能力**:支持 AI Agent 节点,可构建 AI 工作流 + +## Key Features +- 可视化工作流设计 +- 支持 Webhook 触发 +- 内置 AI Agent 模块 +- 通过 n8n-mcp 与 Claude 等 AI 集成 +- 支持 Docker 容器化部署 + +## Related Concepts +- [[n8n-mcp]]:连接 n8n 与 AI 模型的中间件 +- [[n8n节点类型]]:5 种节点分类 +- [[Agentic System]]:n8n 可实现的混合架构 + +## Connections +- [[n8n]] ← 使用 ← [[Telegram]] +- [[n8n]] ← 使用 ← [[Airtable]] +- [[n8n]] ← 使用 ← [[Google Sheets]] +- [[n8n]] ← 运行在 ← [[Docker]] +- [[n8n]] ← 访问 ← [[Caddy]] +- [[Claude]] ← 连接 ← [[n8n]] ← via ← [[n8n-mcp]] \ No newline at end of file diff --git a/wiki/entities/nvm.md b/wiki/entities/nvm.md new file mode 100644 index 00000000..17aab8e1 --- /dev/null +++ b/wiki/entities/nvm.md @@ -0,0 +1,28 @@ +--- +id: nvm +title: "nvm" +type: entity +tags: [node-js, version-manager, tool] +aliases: [Node-Version-Manager] +--- + +## Definition +nvm(Node Version Manager)是一个 Node.js 版本管理器,允许在同一系统上安装和切换多个 Node.js 版本。 + +## Core Properties +- **类型**:工具/版本管理器 +- **安装命令**:`curl -fsSL https://raw.githubusercontent.com/nvm-sh/nvm/v0.39.7/install.sh | bash` +- **官方仓库**:https://github.com/nvm-sh/nvm + +## Usage +```bash +nvm install 20 # 安装 Node 20 +nvm use 20 # 使用 Node 20 +nvm alias default 20 # 设置默认版本 +``` + +## Related Entities +- [[Node-js]]:nvm 管理的 Node.js 运行时 + +## Connections +- [[nvm]] ← 管理 ← [[Node-js]] diff --git a/wiki/entities/v2rayN.md b/wiki/entities/v2rayN.md new file mode 100644 index 00000000..c86823d4 --- /dev/null +++ b/wiki/entities/v2rayN.md @@ -0,0 +1,13 @@ +--- +title: "v2rayN" +type: entity +tags: [代理, 工具] +sources: [an-zhuang-v2rayn.md] +last_updated: 2026-04-14 +--- + +## Definition +跨平台代理工具,支持Windows/macOS/Linux。 + +## Connections +- [[Ubuntu-Server]] ← 工具 ← [[v2rayN]] diff --git a/wiki/entities/万相营造.md b/wiki/entities/万相营造.md new file mode 100644 index 00000000..2c6d4b5d --- /dev/null +++ b/wiki/entities/万相营造.md @@ -0,0 +1,18 @@ +--- +title: "万相营造" +type: entity +tags: [AI视频, 图生视频, AI, 电商] +sources: [14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md] +last_updated: 2026-04-14 +--- + +## Definition +阿里妈妈推出的AI电商营销工具,通过生成式AI帮助商家快速生成创意内容。 + +## Core Attributes +- 类型:AI电商营销工具 +- 厂商:阿里妈妈 +- 特色:高度还原原图,精准理解长文本提示词 + +## Connections +- [[图生视频]] ← 工具 ← [[万相营造]] diff --git a/wiki/entities/乔布斯-Skill.md b/wiki/entities/乔布斯-Skill.md new file mode 100644 index 00000000..cdf1fb57 --- /dev/null +++ b/wiki/entities/乔布斯-Skill.md @@ -0,0 +1,26 @@ +--- +title: "乔布斯.skill" +type: entity +tags: [AI, Skill, 思维顾问] +sources: [bu-tan-ji-shu, yang-xia-ri-ji-5] +last_updated: 2026-04-14 +--- + +## Definition +通过女娲·Skill框架创建的思维顾问AI Skill,提供职业发展和人生选择方面的建议。 + +## 核心观点(来自skill) +1. **工具不值钱,品味值钱**:AI民主化了工具,但品味没有民主化 +2. **做端到端的事**:做一个从iTunes到iPod到iTunes Store的完整体验 +3. **用死亡过滤器**:问"如果今天是最后一天,我还会做今天要做的事吗" + +## 关键引用 +> "一个人用AI做出一个完整的App,比一个100人的团队里当'AI提示词工程师'强一万倍。" + +## Connections +- [[女娲·Skill]] ← 创建框架 ← [[乔布斯.skill]] +- [[苏东坡]] ← 同类 ← [[乔布斯.skill]] + +## Aliases +- 乔布斯 +- Jobs.skill diff --git a/wiki/entities/即梦AI.md b/wiki/entities/即梦AI.md new file mode 100644 index 00000000..e211ccb5 --- /dev/null +++ b/wiki/entities/即梦AI.md @@ -0,0 +1,18 @@ +--- +title: "即梦AI" +type: entity +tags: [AI视频, 图生视频, AI] +sources: [14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md] +last_updated: 2026-04-14 +--- + +## Definition +字节跳动旗下的一站式AI创意创作平台,支持图生视频功能。 + +## Core Attributes +- 类型:一站式AI创意平台 +- 厂商:字节跳动 +- 特色:首尾帧控制,运镜和运动参数自定义 + +## Connections +- [[图生视频]] ← 工具 ← [[即梦AI]] diff --git a/wiki/entities/可灵AI.md b/wiki/entities/可灵AI.md new file mode 100644 index 00000000..c69554bf --- /dev/null +++ b/wiki/entities/可灵AI.md @@ -0,0 +1,18 @@ +--- +title: "可灵AI" +type: entity +tags: [AI视频, 图生视频, AI] +sources: [14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md] +last_updated: 2026-04-14 +--- + +## Definition +快手推出的AI图片和视频创作平台,专注于真实物理规律模拟和人物动作表情生成。 + +## Core Attributes +- 类型:AI视频创作平台 +- 厂商:快手 +- 特色:1080p高清输出,3D时空联合注意力机制 + +## Connections +- [[图生视频]] ← 工具 ← [[可灵AI]] diff --git a/wiki/entities/女娲-Skill.md b/wiki/entities/女娲-Skill.md new file mode 100644 index 00000000..d9a4351d --- /dev/null +++ b/wiki/entities/女娲-Skill.md @@ -0,0 +1,47 @@ +--- +title: "女娲·Skill" +type: entity +tags: [AI, Skill, 框架] +sources: [yang-xia-ri-ji-5] +last_updated: 2026-04-14 +--- + +## Definition +AI Skill创建框架,通过深度调研提炼真实人物的核心思维框架,变成可运行的AI Skill。 + +## 工作流 +``` +用户输入 → 入口分流 + ↓ + Phase 0.5: 创建技能目录 + ↓ + Phase 1: 6个Agent并行采集信息 + (著作/对话/表达DNA/他者视角/决策/时间线) + ↓ + Phase 1.5: 调研Review检查点 + ↓ + Phase 2: 框架提炼 + (心智模型/决策启发式/表达DNA/价值观/诚实边界) + ↓ + Phase 2.5: 提炼确认检查点 + ↓ + Phase 3: Skill构建 + ↓ + Phase 4: 质量验证 + ↓ + Phase 5: 双Agent精炼 + ↓ + 交付: [人名]-perspective/SKILL.md +``` + +## 已创建Skill +- 苏东坡.perspective(https://github.com/ishenwei/openclaw-skills/tree/main/su-dongpo-perspective) +- 乔布斯.skill + +## Connections +- [[苏东坡]] ← 第一个蒸馏对象 ← [[女娲·Skill]] +- [[乔布斯.skill]] ← 同类 ← [[女娲·Skill]] + +## Aliases +- 女娲 +- nuwa-skill diff --git a/wiki/entities/智谱清影.md b/wiki/entities/智谱清影.md new file mode 100644 index 00000000..53f26f9d --- /dev/null +++ b/wiki/entities/智谱清影.md @@ -0,0 +1,18 @@ +--- +title: "智谱清影" +type: entity +tags: [AI视频, 图生视频, AI] +sources: [14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md] +last_updated: 2026-04-14 +--- + +## Definition +智谱AI推出的图生视频AI工具,支持上传图片生成6秒高清视频。 + +## Core Attributes +- 类型:AI视频生成 +- 特色:30秒生成6秒1440×960高清视频 +- 音效:自带CogSound模型生成匹配音效 + +## Connections +- [[图生视频]] ← 工具 ← [[智谱清影]] diff --git a/wiki/entities/比利哥.md b/wiki/entities/比利哥.md new file mode 100644 index 00000000..77514243 --- /dev/null +++ b/wiki/entities/比利哥.md @@ -0,0 +1,33 @@ +--- +title: "比利哥" +type: entity +tags: [人物, AI, 实践者] +sources: [yang-xia-ri-ji-1, yang-xia-ri-ji-2, yang-xia-ri-ji-3, yang-xia-ri-ji-4, yang-xia-ri-ji-5, yang-long-xia-5-tian, bu-tan-ji-shu] +last_updated: 2026-04-14 +--- + +## Definition +AI Agent实践者,"养虾"系列文章作者,研究如何用AI技术真正提高工作效率。 + +## Core Attributes +- 类型:人物/作者 +- 代号:比利哥 +- 研究方向:OpenClaw多Agent协同、记忆系统、笔记持久化 + +## 核心观点 +1. AI Agent的价值是思维升级而非单点能力提升 +2. 把AI变成会自动整理笔记的实习生 +3. 错误只犯一次,第二次就知道怎么做对 + +## 主要项目 +- OpenClaw多Agent系统管理 +- Obsidian+Gitea持久化笔记系统 +- 女娲·Skill蒸馏苏东坡 + +## Connections +- [[OpenClaw]] ← 创建/维护 ← [[比利哥]] +- [[Obsidian]] ← 使用 ← [[比利哥]] + +## Aliases +- 比利哥 +- shenwei(author字段) diff --git a/wiki/entities/海螺AI.md b/wiki/entities/海螺AI.md new file mode 100644 index 00000000..ab3c899e --- /dev/null +++ b/wiki/entities/海螺AI.md @@ -0,0 +1,18 @@ +--- +title: "海螺AI" +type: entity +tags: [AI视频, 图生视频, AI] +sources: [14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md] +last_updated: 2026-04-14 +--- + +## Definition +MiniMax公司推出的AI视频生成工具,主打高主体一致性和连贯性。 + +## Core Attributes +- 类型:AI视频生成 +- 厂商:MiniMax +- 特色:主体一致性高,支持超出图片内容的文本指令 + +## Connections +- [[图生视频]] ← 工具 ← [[海螺AI]] diff --git a/wiki/entities/清华大学.md b/wiki/entities/清华大学.md new file mode 100644 index 00000000..8c6f66f7 --- /dev/null +++ b/wiki/entities/清华大学.md @@ -0,0 +1,16 @@ +--- +title: "清华大学" +type: entity +tags: [大学, AI, 研究] +sources: [deepseek-handbook] +last_updated: 2026-04-14 +--- + +## Definition +中国顶级大学,发布了DeepSeek使用手册。 + +## 关联 +- 元宇宙文化实验室:发布DeepSeek使用手册 + +## Connections +- [[DeepSeek]] ← 手册主题 ← [[清华大学]] diff --git a/wiki/entities/艺映AI.md b/wiki/entities/艺映AI.md new file mode 100644 index 00000000..922ab85e --- /dev/null +++ b/wiki/entities/艺映AI.md @@ -0,0 +1,18 @@ +--- +title: "艺映AI" +type: entity +tags: [AI视频, 图生视频, AI] +sources: [14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md] +last_updated: 2026-04-14 +--- + +## Definition +MewXAI团队推出的多功能AI视频创作工具,支持运动笔刷控制。 + +## Core Attributes +- 类型:AI视频创作 +- 厂商:MewXAI +- 特色:运动笔刷工具,多平台同步 + +## Connections +- [[图生视频]] ← 工具 ← [[艺映AI]] diff --git a/wiki/entities/苏东坡.md b/wiki/entities/苏东坡.md new file mode 100644 index 00000000..6575e22d --- /dev/null +++ b/wiki/entities/苏东坡.md @@ -0,0 +1,31 @@ +--- +title: "苏东坡" +type: entity +tags: [人物, 历史, AI-Skill] +sources: [yang-xia-ri-ji-5] +last_updated: 2026-04-14 +--- + +## Definition +苏轼(1037-1101),北宋文学家,被蒸馏成AI Skill,成为可对话的数字导师。 + +## 核心心智模型 +1. **进退由时,行藏在我**:庙堂与江湖都是正确选项,判断"此刻我能进吗"+"我内心想进吗" +2. **此心安处是吾乡**:故乡是心安之处,逆境反而诞生最重要的作品 +3. **辞达而已**:文章妙在准确传达,不在辞藻堆砌 +4. **以时受力,逆境转化**:时间/困苦可以转化,逆境是创作土壤 +5. **自出新意,不践古人**:学习古人是为了超越古人 +6. **物我相谙,天人合一**:人与自然融为一体 + +## 决策启发式 +- "被浪打下去还能爬起来的人才叫风流" +- "大江东去,浪淘尽,千古风流人物" +- "人生到处知何似,应似飞鸿踏雪泥" + +## Connections +- [[女娲·Skill]] ← 蒸馏框架 ← [[苏东坡]] +- [[乔布斯.skill]] ← 同类 ← [[苏东坡]] + +## Aliases +- 苏轼 +- Su Dongpo diff --git a/wiki/index.md b/wiki/index.md index 647ecb01..b573a823 100644 --- a/wiki/index.md +++ b/wiki/index.md @@ -1,14 +1,312 @@ # Wiki Index -This file is maintained by the LLM. Updated on every ingest. - ## Overview -- [Overview](overview.md) — living synthesis across all sources +- [Overview](overview.md) ## Sources +- [TikTok PM - Python Django 项目](sources/tiktok-pm-python-django-project.md) +- [不谈技术:普通人该怎么在AI时代赚钱?](sources/bu-tan-ji-shu.md) +- [养虾日记1:我用 OpenClaw 管了 28 万张照片](sources/yang-xia-ri-ji-1.md) +- [养虾日记2:让Agent更懂你:OpenClaw + Self-Improving 复盘](sources/yang-xia-ri-ji-2.md) +- [养虾日记3:用 Obsidian + Gitea 为 AI 助手构建持久化笔记系统](sources/yang-xia-ri-ji-3.md) +- [养虾日记4:一次「Context Limit Exceeded」错误排查](sources/yang-xia-ri-ji-4.md) +- [养虾日记5:深夜与苏轼聊AI](sources/yang-xia-ri-ji-5.md) +- [养龙虾5天血泪史:我的AI Agent为什么总失忆?OpenClaw 记忆调试全记录](sources/yang-long-xia-5-tian.md) +- [Cursor 2.0初学者使用指南](sources/cursor-2-0-beginner-guide.md) +- [Git Push 连接重置问题修复](sources/git-push-connection-reset.md) +- [How to get Youtube Channel ID](sources/youtube-channel-id.md) +- [MCP在Cursor中的集成与应用详解](sources/mcp-in-cursor.md) +- [Templater Obsidian Plugin](sources/templater-plugin.md) +- [These 6 Linux apps let you monitor system resources in style](sources/linux-system-monitor.md) +- [Trae远程开发部署指南](sources/trae-remote-dev-guide.md) +- [Useful Prompt Lib](sources/useful-prompt-lib.md) +- [开发经验与项目规范整理文档](sources/dev-experience-standards.md) +- [我的工具集](sources/ai-tools-collection.md) +- [教學 ChatGPT 先做知識整理,再讓 Canva、 Gamma AI 輸出簡報](sources/chatgpt-knowledge-to-canva-gamma.md) +- [清华出的DeepSeek使用手册](sources/deepseek-handbook.md) +- [ChinaTextbook](sources/china-textbook.md) +- [n8n configure telegram trigger](sources/n8n-configure-telegram-trigger.md) +- [n8n docker install & update](sources/n8n-docker-install-update.md) +- [n8n full tutorial building AI agents in 2025 for Beginners!](sources/n8n-full-tutorial-building-ai-agents-in-2025-for-beginners.md) +- [n8n+Claude 通过自然语言自动化工作流](sources/n8n-Claude-通过自然语言自动化工作流.md) +- [使用Claude自动生成N8N工作流的实操教程](sources/使用Claude自动生成N8N工作流的实操教程.md) +- [家庭网络环境概览 2026-04-03](sources/jia-ting-wang-luo-huan-jing-gai-lan-2026-04-03.md) +- [Mac Mini 安装 FRP 0.65.0(ARM64)操作笔记](sources/mac-mini-an-zhuang-frp-0-65-0-arm64-cao-zuo-bi-ji.md) +- [Ubuntu 安装 FRP 0.65.0(x86_64)操作笔记](sources/ubuntu-an-zhuang-frp-0-65-0-x86-64-cao-zuo-bi-ji.md) +- [通过VPS+内网反向代理实现域名访问内网穿透](sources/tong-guo-vps-nei-wang-fan-xiang-dai-li-shi-xian-yu-ming-fang-wen-nei-wang-chuan-tou.md) +- [Synology NAS + Xiaoya Alist + CloudDrive2 + Plex 家庭影视平台](sources/synology-nas-xiaoya-alist-clouddrive2-plex-media-platform.md) +- [在 Ubuntu 安装 Ollama 并运行 Qwen2.5-Coder 7B](sources/zai-ubuntu-an-zhuang-ollama-bing-yun-hang-qwen2-5-coder-7b.md) +- [Mac Mini 服务器配置:防止自动锁屏与睡眠](sources/mac-mini-fu-wu-qi-pei-zhi-fang-zhi-zi-dong-suo-ping-yu-shui-mian.md) +- [家庭监控方案:Prometheus + Grafana + Node Exporter + cAdvisor + Blackbox](sources/jia-ting-jian-kong-fang-an-prometheus-grafana-node-exporter-cadrvisor-blackbox.md) +- [Ubuntu服务器通过rsync实现日常增量备份](sources/ubuntu-fu-wu-qi-tong-guo-rsync-shi-xian-ri-chang-zeng-liang-bei-fen.md) +- [如何用指纹浏览器安全注册并订阅Claude Pro会员全攻略](sources/ru-he-yong-zhi-wen-liu-lan-qi-an-quan-zhu-ce-bing-ding-yue-claude-pro-hui-yuan-quan-gong-lue.md) +- [可自动化、可扩展、AI增强的电商数据采集与处理系统](sources/ke-zi-dong-hua-ke-kuo-zhan-ai-zeng-qiang-de-dian-shang-shu-ju-cai-ji-yu-chu-li-xi-tong.md) +- [Ubuntu用RustDesk远程登录出现不能使用Wayland登录的错误](sources/ubuntu-yong-rustdesk-yuan-cheng-deng-lu-chu-xian-bu-neng-shi-yong-wayland-deng-lu-de-cuo-wu.md) +- [Ubuntu禁用合盖休眠](sources/ubuntu-jin-yong-he-gai-qian-mian.md) +- [在Synology NAS上安装CloudDrive2](sources/zai-synology-nas-shang-an-zhuang-clouddrive2.md) +- [如何传输Docker images并且在另一个Docker安装](sources/ru-he-chuan-shu-docker-images-bing-qie-zai-yi-ge-docker-an-zhuang.md) +- [如何删除旧的废弃的docker container + volume](sources/ru-he-shan-chu-jiu-de-fei-qi-de-docker-container-volume.md) +- [如何判别你的Linux服务器是x64还是ARM64](sources/ru-he-pan-bie-ni-de-linux-fu-wu-qi-shi-x64-huan-shi-arm64.md) +- [如何在Ubuntu Server上通过NFS挂载Synology NAS上的共享文件夹](sources/ru-he-zai-ubuntu-server-shang-tong-guo-nfs-gua-zai-synology-nas-shang-de-gong-xiang-wen-jian-jia.md) +- [如何在Ubuntu Server安装docker & docker compose](sources/ru-he-zai-ubuntu-server-an-zhuang-docker-compose.md) +- [用Docker安装Jellyfin](sources/yong-docker-an-zhuang-jellyfin.md) +- [用Docker安装Portainer](sources/yong-docker-an-zhuang-portainer.md) +- [用Docker安装it-tools](sources/yong-docker-an-zhuang-it-tools.md) +- [用Docker安装transmission](sources/yong-docker-an-zhuang-transmission.md) +- [macOS 创建与解除 Symbolic Link(OpenClaw 目录映射)](sources/macos-chuang-jian-yu-jie-chu-symbolic-link-openclaw-mu-lu-ying-she.md) +- [Install Apache Superset in Docker](sources/install-apache-superset-in-docker.md) +- [Clonezilla对Ubuntu Server进行全盘镜像备份](sources/clonezilla-dui-ubuntu-server-jin-hang-quan-pan-jing-xiang-bei-fen.md) +- [MinIO + Zipline 自托管图床应用安装教程](sources/minio-zipline-zi-cheng-guan-tu-chuang-ying-yong-an-zhuang-jiao-cheng.md) +- [MySQL MariaDB 数据库详细信息](sources/mysql-mariadb-shu-ju-ku-xiang-xi-xin-xi.md) +- [NodeWarden - 把 Bitwarden 搬上 Cloudflare Workers,彻底告别服务器](sources/nodewarden-ba-bitwarden-ban-dao-cloudflare-workers-shang-che-di-gao-bie-fu-wu-qi.md) +- [RAX50 路由器 更新Merlin Clash订阅](sources/rax50-lu-you-qi-geng-xin-merlin-clash-ding-yue.md) +- [Ubuntu 24.04 enable SSH](sources/ubuntu-24-04-enable-ssh.md) +- [Ubuntu Server科学上网](sources/ubuntu-server-ke-xue-shang-wang.md) +- [安装Ubuntu-24.04.2在HP Zbook工作站笔记本上](sources/an-zhuang-ubuntu-24-04-2-zai-hp-zbook-gong-zuo-zhan-bi-ji-ben-shang.md) +- [安装v2rayN](sources/an-zhuang-v2rayn.md) +- [网件RAX50路由器刷梅林固件与科学上网插件安装教程](sources/wang-jian-rax50-lu-you-qi-shua-mei-lin-gu-jian-yu-ke-xue-shang-wang-cha-jian-an-zhuang-jiao-cheng.md) +- [群晖NAS科学上网方法](sources/qun-hui-nas-ke-xue-shang-wang-fang-fa.md) +- [用Docker安装Navidrome](sources/yong-docker-zhong-an-zhuang-navidrome.md) +- [用Docker安装Apache Superset](sources/yong-docker-an-zhuang-apache-superset.md) +- [用Docker安装Homarr](sources/yong-docker-an-zhuang-homarr.md) +- [Best 7 news API data feeds - AI News](sources/best-7-news-api-data-feeds.md) +- [Designing for Agentic AI](sources/designing-for-agentic-ai.md) +- [GitHub 上 5000 人收藏的 Vibe Coding 神级指南](sources/vibe-coding-cn.md) +- [Google 神级生产力工具,所有 GitHub 开源平替都找到了](sources/notebooklm-open-source-alternatives.md) +- [How to Get the RSS Feed For Any YouTube Channel](sources/youtube-rss-feed.md) +- [If you have multiple interests, do not waste the next 2-3 years](sources/multiple-interests-guide.md) +- [LLMs、RAG、AI Agent 三个到底什么区别?](sources/llms-rag-ai-agent-differences.md) +- [Multi-Agent System Reliability](sources/multi-agent-system-reliability.md) +- [14个免费的AI图生视频工具,用AI让图片动起来](sources/14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md) +- [3.2 万人收藏的 Claude Skills,才是 AI 这条路上最值得研究的一套范式!](sources/32wan-ren-shou-cang-de-claude-skills.md) +- [2025 年 11 个神级 AI 开源平替,GitHub 杀疯了。](sources/2025-nai-ge-shen-ji-ai-kai-yuan-ping-ti.md) +- [如何利用Sora接口实现视频自动化生成工作流](sources/ru-he-li-yong-sora-jie-kou-shi-xian-shi-pin-zi-dong-hua-sheng-cheng-gong-zuo-liu.md) +- [我用 Gemini 3 一口气做了 10 个应用,附教程](sources/wo-yong-gemini-3-yi-qi-zuo-liao-10-ge-ying-yong-fu-jiao-cheng.md) +- [文字生成视频网站推荐](sources/zi-sheng-cheng-shi-pin-wang-zhan-tui-jian.md) +- [系统提示词构建原则](sources/xi-tong-ti-shi-ci-gou-jian-yuan-ze.md) +- [详细!离线部署大模型:ollama+deepseek+open-webui安装使用方法及常见问题解决](sources/xiang-xi-ling-shi-bu-shu-da-mo-xing.md) +- [Nano Banana Pro提示词指南](sources/nano-banana-pro-ti-shi-ci-zhi-nan.md) +- [7 ways I use NotebookLM to make my life easier](sources/7-ways-notebooklm-make-life-easier.md) +- [Nano Banana 提示词框架](sources/nano-banana-ti-shi-ci-kuang-jia.md) +- [A Formalization of Recursive Self-Optimizing Generative Systems](sources/formalization-recursive-self-optimizing-generative-systems.md) +- [AI 解决方案专家培训课程](sources/ai-jie-jue-fang-an-zhuan-jia-pei-xun-ke-cheng.md) +- [Never write another prompt](sources/never-write-another-prompt.md) +- [OpenAI ChatGPT 个性化定义](sources/openai-chatgpt-ge-xing-hua-ding-yi.md) +- [The Picture They Paint of You](sources/the-picture-they-paint-of-you.md) +- [Codecrafters build-your-own-x](sources/build-your-own-x.md) +- [RAG从入门到精通系列1:基础RAG](sources/rag-cong-ru-men-dao-jing-tong-xi-lie-1.md) +- [OpenClaw as Desktop Cowork (AionUi)](sources/aionui-cowork-desktop.md) +- [arXiv Paper Reader](sources/arxiv-paper-reader.md) +- [Autonomous Educational Game Development Pipeline](sources/autonomous-game-dev-pipeline.md) +- [Autonomous Project Management with Subagents](sources/autonomous-project-management.md) +- [Multi-Agent Content Factory](sources/content-factory.md) +- [Custom Morning Brief](sources/custom-morning-brief.md) +- [Daily Reddit Digest](sources/daily-reddit-digest.md) +- [Daily YouTube Digest](sources/daily-youtube-digest.md) +- [Dynamic Dashboard with Sub-agent Spawning](sources/dynamic-dashboard.md) +- [AI-Powered Earnings Tracker](sources/earnings-tracker.md) +- [Event Guest Confirmation](sources/event-guest-confirmation.md) +- [Family Calendar Aggregation & Household Assistant](sources/family-calendar-household-assistant.md) +- [Habit Tracker & Accountability Coach](sources/habit-tracker-accountability-coach.md) +- [Health & Symptom Tracker](sources/health-symptom-tracker.md) +- [Inbox De-clutter](sources/inbox-declutter.md) +- [Personal Knowledge Base (RAG)](sources/knowledge-base-rag.md) +- [LaTeX Paper Writing](sources/latex-paper-writing.md) +- [Local CRM Framework with DenchClaw](sources/local-crm-framework.md) +- [Market Research & Product Factory](sources/market-research-product-factory.md) +- [Automated Meeting Notes & Action Items](sources/meeting-notes-action-items.md) +- [Multi-Agent Specialized Team (Solo Founder Setup)](sources/multi-agent-team.md) +- [Multi-Channel Personal Assistant](sources/multi-channel-assistant.md) +- [Multi-Channel AI Customer Service Platform](sources/multi-channel-customer-service.md) +- [Multi-Source Tech News Digest](sources/multi-source-tech-news-digest.md) +- [OpenClaw + n8n Workflow Orchestration](sources/n8n-workflow-orchestration.md) +- [Goal-Driven Autonomous Tasks](sources/overnight-mini-app-builder.md) +- [Personal CRM with Automatic Contact Discovery](sources/personal-crm.md) +- [Phone-Based Personal Assistant](sources/phone-based-personal-assistant.md) +- [Phone Call Notifications](sources/phone-call-notifications.md) +- [Podcast Production Pipeline](sources/podcast-production-pipeline.md) +- [Polymarket Autopilot: Automated Paper Trading](sources/polymarket-autopilot.md) +- [Pre-Build Idea Validator](sources/pre-build-idea-validator.md) +- [Project State Management System: Event-Driven Alternative to Kanban](sources/project-state-management.md) +- [Second Brain](sources/second-brain.md) +- [Self-Healing Home Server & Infrastructure Management](sources/self-healing-home-server.md) +- [Semantic Memory Search](sources/semantic-memory-search.md) +- [Todoist Task Manager: Agent Task Visibility](sources/todoist-task-manager.md) +- [X Account Analysis](sources/x-account-analysis.md) +- [X/Twitter Automation from Chat](sources/x-twitter-automation.md) +- [YouTube Content Pipeline](sources/youtube-content-pipeline.md) ## Entities +- [Dan Koe](entities/Dan-Koe.md) +- [Django](entities/Django.md) +- [MySQL](entities/MySQL.md) +- [TinyMCE](entities/TinyMCE.md) +- [Bright Data](entities/Bright-Data.md) +- [Gunicorn](entities/Gunicorn.md) +- [Nginx](entities/Nginx.md) +- [OpenClaw](entities/OpenClaw.md) +- [Obsidian](entities/Obsidian.md) +- [Gitea](entities/Gitea.md) +- [比利哥](entities/比利哥.md) +- [苏东坡](entities/苏东坡.md) +- [女娲·Skill](entities/女娲-Skill.md) +- [乔布斯.skill](entities/乔布斯-Skill.md) +- [MiniMax-M2.7](entities/MiniMax-M2.7.md) +- [LanceDB](entities/LanceDB.md) +- [LangChain](entities/LangChain.md) +- [LlamaIndex](entities/LlamaIndex.md) +- [deepseek-reasoner](entities/deepseek-reasoner.md) +- [Karpathy](entities/Karpathy.md) +- [QMD](entities/QMD.md) +- [Qdrant](entities/Qdrant.md) +- [RAG](entities/RAG.md) +- [Cursor](entities/Cursor.md) +- [Trae](entities/Trae.md) +- [GitHub](entities/GitHub.md) +- [YouTube](entities/YouTube.md) +- [DeepSeek](entities/DeepSeek.md) +- [Flux](entities/Flux.md) +- [清华大学](entities/清华大学.md) +- [ChatGPT](entities/ChatGPT.md) +- [Canva](entities/Canva.md) +- [Gamma](entities/Gamma.md) +- [n8n](entities/n8n.md) +- [NotebookLM](entities/NotebookLM.md) +- [ChinaTextbook](entities/ChinaTextbook.md) +- [Templater](entities/Templater.md) +- [Docker](entities/Docker.md) +- [btop++](entities/b_top_plus_plus.md) +- [Telegram](entities/Telegram.md) +- [Cpolar](entities/Cpolar.md) +- [Airtable](entities/Airtable.md) +- [Node.js](entities/Node_js.md) +- [Caddy](entities/Caddy.md) +- [Cline](entities/Cline.md) +- [Coze](entities/Coze.md) +- [Dify](entities/Dify.md) +- [V2Ray/Tuic](entities/V2Ray-Tuic.md) +- [Google Sheets](entities/Google-Sheets.md) +- [Mac Mini](entities/Mac-Mini.md) +- [Manus](entities/Manus.md) +- [VPS](entities/VPS.md) +- [Synology NAS](entities/Synology-NAS.md) +- [Ubuntu Server](entities/Ubuntu-Server.md) +- [Plex](entities/Plex.md) +- [Perplexica](entities/Perplexica.md) +- [Ollama](entities/Ollama.md) +- [RustDesk](entities/RustDesk.md) +- [CloudDrive2](entities/CloudDrive2.md) +- [Jellyfin](entities/Jellyfin.md) +- [Portainer](entities/Portainer.md) +- [it-tools](entities/it-tools.md) +- [Transmission](entities/Transmission.md) +- [Clonezilla](entities/Clonezilla.md) +- [Rufus](entities/Rufus.md) +- [Apache Superset](entities/Apache-Superset.md) +- [NodeWarden](entities/NodeWarden.md) +- [Zipline](entities/Zipline.md) +- [MinIO](entities/MinIO.md) +- [Merlin Clash](entities/Merlin-Clash.md) +- [RAX50](entities/RAX50.md) +- [OpenSSH](entities/OpenSSH.md) +- [OpenManus](entities/OpenManus.md) +- [Navidrome](entities/Navidrome.md) +- [Homarr](entities/Homarr.md) +- [Anthropic](entities/Anthropic.md) +- [Claude Skills](entities/Claude-Skills.md) +- [v2rayN](entities/v2rayN.md) +- [Sora](entities/Sora.md) +- [Gemini-3](entities/Gemini-3.md) +- [Nano Banana Pro](entities/Nano-Banana-Pro.md) +- [Vidu](entities/Vidu.md) +- [Video Ocean](entities/Video-Ocean.md) +- [Viva](entities/Viva.md) +- [万相营造](entities/万相营造.md) +- [即梦AI](entities/即梦AI.md) +- [可灵AI](entities/可灵AI.md) +- [海螺AI](entities/海螺AI.md) +- [艺映AI](entities/艺映AI.md) +- [智谱清影](entities/智谱清影.md) +- [PixVerse](entities/PixVerse.md) +- [Stable Video](entities/Stable-Video.md) +- [Stable Diffusion](entities/Stable-Diffusion.md) +- [Haiper](entities/Haiper.md) +- [HunyuanVideo](entities/HunyuanVideo.md) ## Concepts +- [Django Admin定制](concepts/Django-Admin定制.md) +- [RESTful API设计](concepts/RESTful-API设计.md) +- [Docker容器化部署](concepts/Docker容器化部署.md) +- [异步任务处理](concepts/异步任务处理.md) +- [Self-Improving Skill](concepts/Self-Improving-Skill.md) +- [LLM Wiki](concepts/LLM-Wiki.md) +- [三层记忆架构](concepts/三层记忆架构.md) +- [Context Window](concepts/Context-Window.md) +- [Compaction(内存压缩)](concepts/Compaction.md) +- [混合搜索](concepts/混合搜索.md) +- [Skill蒸馏](concepts/Skill蒸馏.md) +- [死亡过滤器](concepts/死亡过滤器.md) +- [端到端能力](concepts/端到端能力.md) +- [品味值链](concepts/品味值链.md) +- [认知操作系统](concepts/认知操作系统.md) +- [写入纪律](concepts/写入纪律.md) +- [交接协议](concepts/交接协议.md) +- [Memory Flush](concepts/Memory-Flush.md) +- [MCP](concepts/MCP.md) +- [Agent模式](concepts/Agent模式.md) +- [Composer](concepts/Composer.md) +- [Git代理配置](concepts/Git代理配置.md) +- [SSH协议切换](concepts/SSH协议切换.md) +- [远程开发](concepts/远程开发.md) +- [变量命名规范](concepts/变量命名规范.md) +- [单一职责](concepts/单一职责.md) +- [DRY原则](concepts/DRY原则.md) +- [知识整理先行](concepts/知识整理先行.md) +- [防弹笔记法](concepts/防弹笔记法.md) +- [TUI](concepts/TUI.md) +- [Agentic System](concepts/Agentic-System.md) +- [n8n节点类型](concepts/n8n节点类型.md) +- [自然语言工作流生成](concepts/自然语言工作流生成.md) +- [Extended Thinking](concepts/Extended-Thinking.md) +- [Opensea模型](concepts/Opensea模型.md) +- [n8n-mcp](concepts/n8n-mcp.md) +- [内网穿透](concepts/内网穿透.md) +- [Caddy](concepts/Caddy.md) +- [FRP](concepts/FRP.md) +- [云盘挂载](concepts/云盘挂载.md) +- [监控体系](concepts/监控体系.md) +- [增量备份](concepts/增量备份.md) +- [指纹浏览器](concepts/指纹浏览器.md) +- [本地LLM](concepts/本地LLM.md) +- [Symbolic Link](concepts/Symbolic-Link.md) +- [磁盘镜像备份](concepts/磁盘镜像备份.md) +- [S3兼容存储](concepts/S3兼容存储.md) +- [图片托管服务](concepts/图片托管服务.md) +- [逻辑备份](concepts/逻辑备份.md) +- [无服务器密码管理](concepts/无服务器密码管理.md) +- [Cloudflare D1](concepts/Cloudflare-D1.md) +- [Cloudflare R2](concepts/Cloudflare-R2.md) +- [路由器科学上网](concepts/路由器科学上网.md) +- [Socket激活机制](concepts/Socket激活机制.md) +- [终端代理](concepts/终端代理.md) +- [Docker代理配置](concepts/Docker代理配置.md) +- [视频自动化生成工作流](concepts/视频自动化生成工作流.md) +- [提示词优化](concepts/提示词优化.md) +- [提示词框架](concepts/提示词框架.md) +- [Gemini-3应用开发方法论](concepts/Gemini-3应用开发方法论.md) +- [文字生成视频](concepts/文字生成视频.md) +- [系统提示词构建](concepts/系统提示词构建.md) +- [身份锁定](concepts/身份锁定.md) +- [意图理解引擎](concepts/意图理解引擎.md) +- [图生视频](concepts/图生视频.md) +- [Claude Skills](concepts/Claude-Skills.md) +- [流程工程](concepts/流程工程.md) +- [递归自优化](concepts/递归自优化.md) +- [固定点理论](concepts/固定点理论.md) +- [Embedding Vector](concepts/Embedding-Vector.md) +- [Generation](concepts/Generation.md) +- [Indexing](concepts/Indexing.md) +- [Retrieval](concepts/Retrieval.md) +- [Vector Store](concepts/Vector-Store.md) ## Syntheses diff --git a/wiki/log.md b/wiki/log.md index 66a92854..53ff957f 100644 --- a/wiki/log.md +++ b/wiki/log.md @@ -1,9 +1,124 @@ # Wiki Log -Append-only chronological record of all operations. - -Format: `## [YYYY-MM-DD] <operation> | <title>` - -Parse recent entries: `grep "^## \[" wiki/log.md | tail -10` - ---- +## [2026-04-14] ingest | TikTok PM - Python Django 项目 +## [2026-04-14] ingest | 不谈技术:普通人该怎么在AI时代赚钱? +## [2026-04-14] ingest | 养虾日记1:我用 OpenClaw 管了 28 万张照片 +## [2026-04-14] ingest | 养虾日记2:让Agent更懂你:OpenClaw + Self-Improving 复盘 +## [2026-04-14] ingest | 养虾日记3:用 Obsidian + Gitea 为 AI 助手构建持久化笔记系统 +## [2026-04-14] ingest | 养虾日记4:一次「Context Limit Exceeded」错误排查 +## [2026-04-14] ingest | 养虾日记5:深夜与苏轼聊AI +## [2026-04-14] ingest | 养龙虾5天血泪史:我的AI Agent为什么总失忆? +## [2026-04-14] ingest | Cursor 2.0初学者使用指南 +## [2026-04-14] ingest | Git Push 连接重置问题修复 +## [2026-04-14] ingest | How to get Youtube Channel ID +## [2026-04-14] ingest | MCP在Cursor中的集成与应用详解 +## [2026-04-14] ingest | Templater Obsidian Plugin +## [2026-04-14] ingest | These 6 Linux apps let you monitor system resources in style +## [2026-04-14] ingest | Trae远程开发部署指南 +## [2026-04-14] ingest | Useful Prompt Lib +## [2026-04-14] ingest | 开发经验与项目规范整理文档 +## [2026-04-14] ingest | 我的工具集 +## [2026-04-14] ingest | 教學 ChatGPT 先做知識整理,再讓 Canva、 Gamma AI 輸出簡報 +## [2026-04-14] ingest | 清华出的DeepSeek使用手册 +## [2026-04-14] ingest | ChinaTextbook +## [2026-04-14] ingest | n8n configure telegram trigger +## [2026-04-14] ingest | n8n docker install & update +## [2026-04-14] ingest | n8n full tutorial building AI agents in 2025 for Beginners! +## [2026-04-14] ingest | n8n+Claude 通过自然语言自动化工作流 +## [2026-04-14] ingest | 使用Claude自动生成N8N工作流的实操教程 +## [2026-04-14] ingest | 家庭网络环境概览 2026-04-03 +## [2026-04-14] ingest | Mac Mini 安装 FRP 0.65.0(ARM64)操作笔记 +## [2026-04-14] ingest | Ubuntu 安装 FRP 0.65.0(x86_64)操作笔记 +## [2026-04-14] ingest | 通过VPS+内网反向代理实现域名访问内网穿透 +## [2026-04-14] ingest | Synology NAS + Xiaoya Alist + CloudDrive2 + Plex 家庭影视平台 +## [2026-04-14] ingest | 在 Ubuntu 安装 Ollama 并运行 Qwen2.5-Coder 7B +## [2026-04-14] ingest | Mac Mini 服务器配置:防止自动锁屏与睡眠 +## [2026-04-14] ingest | 家庭监控方案:Prometheus + Grafana + Node Exporter + cAdvisor + Blackbox +## [2026-04-14] ingest | Ubuntu服务器通过rsync实现日常增量备份 +## [2026-04-14] ingest | 如何用指纹浏览器安全注册并订阅Claude Pro会员全攻略 +## [2026-04-14] ingest | 可自动化、可扩展、AI增强的电商数据采集与处理系统 +## [2026-04-14] ingest | Ubuntu用RustDesk远程登录出现不能使用Wayland登录的错误 +## [2026-04-14] ingest | Ubuntu禁用合盖休眠 +## [2026-04-14] ingest | 在Synology NAS上安装CloudDrive2 +## [2026-04-14] ingest | 如何传输Docker images并且在另一个Docker安装 +## [2026-04-14] ingest | 如何删除旧的废弃的docker container + volume +## [2026-04-14] ingest | 如何判别你的Linux服务器是x64还是ARM64 +## [2026-04-14] ingest | 如何在Ubuntu Server上通过NFS挂载Synology NAS上的共享文件夹 +## [2026-04-14] ingest | 如何在Ubuntu Server安装docker & docker compose +## [2026-04-14] ingest | 用Docker安装Jellyfin +## [2026-04-14] ingest | 用Docker安装Portainer +## [2026-04-14] ingest | 用Docker安装it-tools +## [2026-04-14] ingest | 用Docker安装transmission +## [2026-04-14] ingest | macOS 创建与解除 Symbolic Link(OpenClaw 目录映射) +## [2026-04-14] ingest | Install Apache Superset in Docker +## [2026-04-14] ingest | Clonezilla对Ubuntu Server进行全盘镜像备份 +## [2026-04-14] ingest | MinIO + Zipline 自托管图床应用安装教程 +## [2026-04-14] ingest | MySQL MariaDB 数据库详细信息 +## [2026-04-14] ingest | NodeWarden - 把 Bitwarden 搬上 Cloudflare Workers,彻底告别服务器 +## [2026-04-14] ingest | RAX50 路由器 更新Merlin Clash订阅 +## [2026-04-14] ingest | Ubuntu 24.04 enable SSH +## [2026-04-14] ingest | Ubuntu Server科学上网 +## [2026-04-14] ingest | 安装Ubuntu-24.04.2在HP Zbook工作站笔记本上 +## [2026-04-14] ingest | 安装v2rayN +## [2026-04-14] ingest | 网件RAX50路由器刷梅林固件与科学上网插件安装教程 +## [2026-04-14] ingest | 群晖NAS科学上网方法 +## [2026-04-14] ingest | 用Docker安装Navidrome +## [2026-04-14] ingest | 用Docker安装Apache Superset +## [2026-04-14] ingest | 用Docker安装Homarr +## [2026-04-14] ingest | 如何利用Sora接口实现视频自动化生成工作流 +## [2026-04-14] ingest | 我用 Gemini 3 一口气做了 10 个应用,附教程 +## [2026-04-14] ingest | 文字生成视频网站推荐 +## [2026-04-14] ingest | 系统提示词构建原则 +## [2026-04-14] ingest | 详细!离线部署大模型:ollama+deepseek+open-webui安装使用方法及常见问题解决 +## [2026-04-14] ingest | Nano Banana Pro提示词指南 +## [2026-04-14] ingest | 14个免费的AI图生视频工具,用AI让图片动起来 +## [2026-04-14] ingest | 2025 年 11 个神级 AI 开源平替,GitHub 杀疯了。 +## [2026-04-14] ingest | 7 ways I use NotebookLM to make my life easier +## [2026-04-14] ingest | Nano Banana 提示词框架 +## [2026-04-14] ingest | A Formalization of Recursive Self-Optimizing Generative Systems +## [2026-04-14] ingest | AI 解决方案专家培训课程 +## [2026-04-14] ingest | Never write another prompt +## [2026-04-14] ingest | OpenAI ChatGPT 个性化定义 +## [2026-04-14] ingest | The Picture They Paint of You +## [2026-04-14] ingest | Codecrafters build-your-own-x +## [2026-04-14] ingest | RAG从入门到精通系列1:基础RAG +## [2026-04-14] ingest | OpenClaw as Desktop Cowork (AionUi) +## [2026-04-14] ingest | arXiv Paper Reader +## [2026-04-14] ingest | Autonomous Educational Game Development Pipeline +## [2026-04-14] ingest | Autonomous Project Management with Subagents +## [2026-04-14] ingest | Multi-Agent Content Factory +## [2026-04-14] ingest | Custom Morning Brief +## [2026-04-14] ingest | Daily Reddit Digest +## [2026-04-14] ingest | Daily YouTube Digest +## [2026-04-14] ingest | Dynamic Dashboard with Sub-agent Spawning +## [2026-04-14] ingest | AI-Powered Earnings Tracker +## [2026-04-14] ingest | Event Guest Confirmation +## [2026-04-14] ingest | Family Calendar Aggregation & Household Assistant +## [2026-04-14] ingest | Habit Tracker & Accountability Coach +## [2026-04-14] ingest | Health & Symptom Tracker +## [2026-04-14] ingest | Inbox De-clutter +## [2026-04-14] ingest | Personal Knowledge Base (RAG) +## [2026-04-14] ingest | LaTeX Paper Writing +## [2026-04-14] ingest | Local CRM Framework with DenchClaw +## [2026-04-14] ingest | Market Research & Product Factory +## [2026-04-14] ingest | Automated Meeting Notes & Action Items +## [2026-04-14] ingest | Multi-Agent Specialized Team (Solo Founder Setup) +## [2026-04-14] ingest | Multi-Channel Personal Assistant +## [2026-04-14] ingest | Multi-Channel AI Customer Service Platform +## [2026-04-14] ingest | Multi-Source Tech News Digest +## [2026-04-14] ingest | OpenClaw + n8n Workflow Orchestration +## [2026-04-14] ingest | Goal-Driven Autonomous Tasks +## [2026-04-14] ingest | Personal CRM with Automatic Contact Discovery +## [2026-04-14] ingest | Phone-Based Personal Assistant +## [2026-04-14] ingest | Phone Call Notifications +## [2026-04-14] ingest | Podcast Production Pipeline +## [2026-04-14] ingest | Polymarket Autopilot: Automated Paper Trading +## [2026-04-14] ingest | Pre-Build Idea Validator +## [2026-04-14] ingest | Project State Management System +## [2026-04-14] ingest | Second Brain +## [2026-04-14] ingest | Self-Healing Home Server & Infrastructure Management +## [2026-04-14] ingest | Semantic Memory Search +## [2026-04-14] ingest | Todoist Task Manager: Agent Task Visibility +## [2026-04-14] ingest | X Account Analysis +## [2026-04-14] ingest | X/Twitter Automation from Chat +## [2026-04-14] ingest | YouTube Content Pipeline diff --git a/wiki/overview.md b/wiki/overview.md index f71416d2..23a7beb4 100644 --- a/wiki/overview.md +++ b/wiki/overview.md @@ -1,17 +1,65 @@ ---- -title: "Overview" -type: synthesis -tags: [] -sources: [] -last_updated: "" ---- +# Wiki Overview -# Overview +## 主题 +AI Agent实践与知识管理——以OpenClaw为核心的Personal AI OS构建 -*This page is maintained by the LLM. It is updated on every ingest to reflect the current synthesis across all sources.* +## 核心实体 +- **OpenClaw**:AI Agent操作系统,管理多个专业化Agent(星枢、星辉、星曜) +- **Cursor/Trae**:AI增强代码编辑器,支持多代理和远程开发 +- **比利哥**:实践者,养虾(研究AI Agent应用)系列文章作者 +- **Obsidian + Gitea**:持久化笔记基础设施 +- **女娲·Skill**:Skill蒸馏框架 +- **ChatGPT + Canva/Gamma**:AI简报制作工作流 -No sources ingested yet. Add your first source with: +## 核心概念 -```bash -python tools/ingest.py raw/your-source.md -``` +### AI开发工具 +- **MCP协议**:Modal Context Protocol,AI大模型与外围工具的集成交互 +- **Cursor Composer**:自研AI模型,生成速度比同类快4倍 +- **Agent模式**:自动执行命令,与Normal/Ask模式区别 + +### 记忆系统 +三层记忆架构解决Agent失忆问题: +1. **短期记忆**:memory/YYYY-MM-DD.md(每日对话文件) +2. **长期记忆**:LanceDB向量数据库 +3. **Self-Improving**:结构化复盘,错误只犯一次 + +### 持久化 +- AI输出必须写盘才能在Context压缩中存活 +- 交接协议解决模型切换时上下文丢失 +- 写入纪律比读取纪律更重要 + +### 知识管理 +- **LLM Wiki vs RAG**:RAG从零检索,LLM Wiki增量累积 +- **Graph View**:发现孤岛页面和知识盲区 +- **混合搜索**:BM25 + 向量 + 重排序 +- **知识整理先行**:简报不是从版面设计开始,而是从资料研究开始 + +### 开发规范 +- **变量命名规范**:小写英文+下划线或小驼峰 +- **单一职责**:每个文件/类/函数只负责一件事 +- **DRY原则**:Don't Repeat Yourself + +### 远程开发 +- **Trae Remote-SSH**:通过SSH远程连接开发服务器 +- **Docker容器化**:开发/生产环境分离 +- **Git代理配置**:解决国内访问GitHub的连接问题 + +### 思维方式 +- **死亡过滤器**:对一千件事说No,只对一件事说Yes +- **端到端能力**:做完整的产品,不做零件 +- **品味值链**:AI工具民主化,品味是护城河 + +## 关键教训 +1. 自动加载的文件只有7个:AGENTS.md/SOUL.md/TOOLS.md/IDENTITY.md/USER.md/HEARTBEAT.md/MEMORY.md +2. 启动序列必须放在AGENTS.md顶部 +3. 压缩不是敌人,未写入的上下文才是 +4. 错误信息可能是表面意思,真正问题藏在配置层级深处 +5. 真正的修复不是添加更多文件,而是移除什么都不做的文件 +6. 简报不是从版面设计开始,而是从资料研究开始 + +## 知识空白 +- Self-Improving的Pattern-Key设计可进一步优化 +- LLM Wiki与RAG的边界条件尚未明确 +- Skill蒸馏的质量验证流程可自动化 +- MCP工具链的标准化集成流程 diff --git a/wiki/sources/14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md b/wiki/sources/14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md new file mode 100644 index 00000000..01c35933 --- /dev/null +++ b/wiki/sources/14ge-mian-fei-ai-tu-sheng-shi-ping-gong-ju.md @@ -0,0 +1,60 @@ +--- +title: "14个免费的AI图生视频工具,用AI让图片动起来" +type: source +tags: [AI, image-to-video, AI视频] +sources: ["https://www.51juzd.com/23332.html"] +date: 2025-12-05 +--- + +## Summary +- 核心主题:14款免费AI图生视频工具对比与使用指南 +- 问题域:创作者如何利用免费AI工具将静态图片转化为动态视频 +- 方法/机制:上传静态图片+选择动作模板/输入提示词→AI生成动态视频 +- 结论/价值:为内容创作者提供免费图生视频工具的全面的功能对比和选型参考 + +## Key Claims +- 14款工具覆盖中国头部AI公司和海外知名平台,免费为主 +- 主要功能差异:运动控制精度、风格多样性、分辨率、视频时长、主体一致性 +- 中国平台(绘蛙、智谱清影、通义万相、Vidu、可灵、海螺、即梦)侧重电商和内容创作场景 +- 海外平台(Stable Video、Viva、PixVerse)侧重创作者工作流 + +## Key Concepts +- [[图生视频]](Image-to-Video):将静态图片转化为动态视频的AI生成技术 +- [[主体一致性]]:视频中人物/物体保持与原图一致的技术能力 +- [[运镜控制]]:通过参数控制视频镜头运动方式 +- [[提示词控制]]:通过文本描述控制视频中主体的运动和场景变化 + +## Key Entities +- [[绘蛙AI视频]]:阿里巴巴模特图生视频工具 +- [[智谱清影]]:智谱AI视频生成,30秒生成6秒1440×960视频 +- [[通义万相]]:阿里巴巴AI视频生成,支持音效匹配 +- [[Vidu]]:生数科技+清华大学,长时长高一致性视频模型 +- [[可灵AI]]:快手AI视频,物理规律模拟能力强 +- [[海螺AI]]:MiniMax公司视频生成,主体一致性高 +- [[即梦AI]]:字节跳动一站式AI创意平台 +- [[PixVerse]]:爱诗科技视频生成,支持首尾帧 +- [[Video Ocean]]:潞晨科技视频生成,多种画风 +- [[Stable Video]]:Stability AI视频平台,摄像机精细控制 +- [[万相营造]]:阿里妈妈电商营销工具 +- [[Viva]]:智象未来免费AI创意平台 +- [[Haiper]]:AI视频生成,Discord免费使用 +- [[艺映AI]]:MewXAI团队,支持运动笔刷 + +## Connections +- [[图生视频]] ← 工具 ← [[绘蛙AI视频]] +- [[图生视频]] ← 工具 ← [[智谱清影]] +- [[图生视频]] ← 工具 ← [[通义万相]] +- [[图生视频]] ← 工具 ← [[Vidu]] +- [[图生视频]] ← 工具 ← [[可灵AI]] +- [[图生视频]] ← 工具 ← [[海螺AI]] +- [[图生视频]] ← 工具 ← [[即梦AI]] +- [[图生视频]] ← 工具 ← [[PixVerse]] +- [[图生视频]] ← 工具 ← [[Video Ocean]] +- [[图生视频]] ← 工具 ← [[Stable Video]] +- [[图生视频]] ← 工具 ← [[万相营造]] +- [[图生视频]] ← 工具 ← [[Viva]] +- [[图生视频]] ← 工具 ← [[Haiper]] +- [[图生视频]] ← 工具 ← [[艺映AI]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/2025-nai-ge-shen-ji-ai-kai-yuan-ping-ti.md b/wiki/sources/2025-nai-ge-shen-ji-ai-kai-yuan-ping-ti.md new file mode 100644 index 00000000..7da539e4 --- /dev/null +++ b/wiki/sources/2025-nai-ge-shen-ji-ai-kai-yuan-ping-ti.md @@ -0,0 +1,52 @@ +--- +title: "2025 年 11 个神级 AI 开源平替,GitHub 杀疯了。" +type: source +tags: [AI开源, GitHub, LLM, AI绘图, AI视频, AI智能体] +sources: ["https://mp.weixin.qq.com/s/nEXgzvE2FUGBXCHkmbWifg"] +date: 2026-01-01 +--- + +## Summary +- 核心主题:2025年11个AI细分领域的GitHub最强开源平替汇总 +- 问题域:各AI产品闭源替代方案选型 +- 方法/机制:按领域分类介绍Star最高、功能最强的开源项目 +- 结论/价值:为用户提供各AI领域的开源替代完整图谱 + +## Key Claims +- LLM领域:DeepSeek R1是首个将o1级深度推理拉下神坛的开源破壁者;Qwen 3是最稳最全的六边形战士 +- AI生图:Flux是目前人体解剖学最正确的开源模型;SD 3.5的LoRA和ControlNet生态最丰富 +- AI生视频:HunyuanVideo是开源界参数量最大的视频生成模型之一,中文理解能力强 +- AI智能体:OpenManus有5万Star,核心逻辑是规划→执行→循环反馈 +- AI Coding:Cline是Cursor的最佳开源平替,支持MCP扩展和用户授权机制 +- 工作流:n8n(16万Star)和Dify是最强的Workflow开源项目 +- AI搜索:Perplexica可本地化部署,支持SearXNG和本地LLM + +## Key Entities +- [[DeepSeek]]:开源R1和V3,深度推理能力强 +- [[Qwen]]:通义千问,开源六边形战士 +- [[Flux]]:开源界Midjourney,人体解剖学最正确 +- [[HunyuanVideo]]:腾讯混元视频,参数量最大 +- [[OpenManus]]:5万Star,规划→执行→循环反馈 +- [[Cline]]:Cursor最佳开源平替,VS Code插件 +- [[n8n]]:16万Star,工作流开源项目 +- [[Dify]]:LLM应用开发平台 +- [[Perplexica]]:Perplexity开源平替,2.8K Star +- [[Manus]]:被Meta数十亿收购,定义AI Agent元年 +- [[Nano Banana]]:AI绘图工具 +- [[Cline]]:AI编程工具 + +## Connections +- [[AI开源平替]] ← LLM ← [[DeepSeek]] +- [[AI开源平替]] ← LLM ← [[Qwen]] +- [[AI开源平替]] ← 生图 ← [[Flux]] +- [[AI开源平替]] ← 生图 ← [[Stable-Diffusion]] +- [[AI开源平替]] ← 生视频 ← [[HunyuanVideo]] +- [[AI开源平替]] ← 智能体 ← [[OpenManus]] +- [[AI开源平替]] ← 智能体 ← [[Manus]] +- [[AI开源平替]] ← Coding ← [[Cline]] +- [[AI开源平替]] ← 工作流 ← [[n8n]] +- [[AI开源平替]] ← 工作流 ← [[Dify]] +- [[AI开源平替]] ← 搜索 ← [[Perplexica]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/32wan-ren-shou-cang-de-claude-skills.md b/wiki/sources/32wan-ren-shou-cang-de-claude-skills.md new file mode 100644 index 00000000..d696dee4 --- /dev/null +++ b/wiki/sources/32wan-ren-shou-cang-de-claude-skills.md @@ -0,0 +1,49 @@ +--- +title: "3.2 万人收藏的 Claude Skills,才是 AI 这条路上最值得研究的一套范式!" +type: source +tags: [claude-skills, ai, workflow, vibe-coding] +sources: ["https://mp.weixin.qq.com/s/eBAt1OBPZVobyZlcuNPeAw"] +date: 2026-01-05 +--- + +## Summary +- 核心主题:Claude Skills(技能)范式解析与资源汇总 +- 问题域:如何从提示词工程升级到流程工程 +- 方法/机制:Skills = 说明书 + SOP,将重复任务拆解为AI可理解、可复用、可自动执行的流程 +- 结论/价值:Skills标志着从"提示词工程"到"流程工程"的质变,是AI应用逻辑的升级 + +## Key Claims +- Skills是Anthropic官方发布的AI技能指南,本质是写给Claude的"说明书"和"SOP" +- 官方Skills仓库展示了真实生产级能力:Office四大件、开发者工具箱、创意类Skill +- Skills的核心价值:可复用、可自动执行、有边界 +- 三大Skill聚合站:skillsmp.com、aitmpl.com、claudemarketplaces.com +- 未来真正有价值的是:谁最懂业务流程、谁能将经验沉淀成SOP + +## Key Concepts +- [[Claude Skills]](技能):写给AI的说明书和SOP,将重复任务拆解为可复用流程 +- [[SOP]](标准作业程序):Skills的核心组成部分 +- [[流程工程]]:从提示词工程升级后的新范式 +- [[Vibe Coding]] ← 尽头是 ← [[Claude Skills]] + +## Key Entities +- [[Anthropic]]:Claude Skills的发布方 +- [[Claude Skills官方仓库]]:https://github.com/anthropics/skills,3.2万星 +- [[awesome-claude-skills]]:ComposioHQ/VoltAgent/BehiSecc三个精选仓库 +- [[skillsmp.com]]:Skill聚合站 +- [[aitmpl.com]]:Skill聚合站 +- [[claudemarketplaces.com]]:Skill聚合站 + +## Connections +- [[Claude Skills]] ← 发布方 ← [[Anthropic]] +- [[Claude Skills]] ← 包含 ← [[SOP]] +- [[Claude Skills]] ← 升级自 ← [[提示词工程]] +- [[Vibe Coding]] ← 尽头 ← [[Claude Skills]] +- [[Claude Skills]] ← 聚合 ← [[skillsmp.com]] +- [[Claude Skills]] ← 聚合 ← [[aitmpl.com]] +- [[Claude Skills]] ← 聚合 ← [[claudemarketplaces.com]] + +## Contradictions +- 与纯提示词工程观点冲突: + - 冲突点:提示词 vs SOP + - 当前观点:Skills是可复用的SOP,远优于提示词 + - 对方观点:优化提示词即可实现AI应用 diff --git a/wiki/sources/7-ways-notebooklm-make-life-easier.md b/wiki/sources/7-ways-notebooklm-make-life-easier.md new file mode 100644 index 00000000..ca5b5bfa --- /dev/null +++ b/wiki/sources/7-ways-notebooklm-make-life-easier.md @@ -0,0 +1,39 @@ +--- +title: "7 ways I use NotebookLM to make my life easier" +type: source +tags: [notebooklm, productivity, AI工具, 学习] +sources: ["https://www.howtogeek.com/ways-notebooklm-make-my-life-easier/"] +date: 2025-11-23 +--- + +## Summary +- 核心主题:NotebookLM的7种实际应用场景 +- 问题域:如何利用NotebookLM提升日常效率 +- 方法/机制:source-grounding确保答案准确,可上传多种格式文档,AI自动处理并生成结构化输出 +- 结论/价值:NotebookLM不仅是研究工具,更可作为个人项目管理和文档分析的AI助手 + +## Key Claims +- source-grounding是核心机制:NotebookLM的知识库严格限制在用户上传的文档内,保证输出准确可验证 +- Audio Overviews将文档转化为双人对话播客,支持自定义提示词控制风格和焦点 +- 可作为项目管理和目标规划工具,将分散的研究和想法整合为清晰的路线图 +- 适合法律文档分析:每个答案都有精确引用,可点击查看原文 +- 学习编程更快:将文档导入Notebook,用问答代替看长篇教程 + +## Key Concepts +- [[Source-grounding]]:NotebookLM的核心机制,严格限制知识库在用户上传文档内 +- [[Audio Overviews]]:NotebookLM的播客功能,将文档转化为双人AI对话音频 +- [[被动学习]]:利用通勤、健身等时间通过音频消费复杂信息 + +## Key Entities +- [[NotebookLM]]:Google AI笔记助手,source-grounding确保准确性 + +## Connections +- [[NotebookLM]] ← 应用场景 ← [[信息处理]] +- [[NotebookLM]] ← 应用场景 ← [[被动学习]] +- [[NotebookLM]] ← 应用场景 ← [[编程学习]] +- [[NotebookLM]] ← 应用场景 ← [[项目管理]] +- [[NotebookLM]] ← 应用场景 ← [[版本对比]] +- [[NotebookLM]] ← 应用场景 ← [[文档分析]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/agentic-ai-cloud-devops.md b/wiki/sources/agentic-ai-cloud-devops.md new file mode 100644 index 00000000..5a6b879c --- /dev/null +++ b/wiki/sources/agentic-ai-cloud-devops.md @@ -0,0 +1,39 @@ +--- +title: "How Agentic AI can help for Cloud DevOps" +type: source +tags: [Agentic AI, Cloud, DevOps, Automation] +sources: [] +last_updated: 2025-03-01 +--- + +## Summary +- 核心主题:Agentic AI(自主决策AI系统)如何增强云DevOps能力 +- 问题域:利用AI自动化复杂工作流、提升效率、确保云环境可靠性 +- 方法/机制:七大AI增强领域(事件检测与解决、自动部署、成本优化、安全合规、日志分析、多租户管理、AI辅助决策) +- 结论/价值:Agentic AI通过自动化事件响应、成本管理、安全、可观测性和多云治理转变云DevOps + +## Key Claims +- Agentic AI可实现自主事件检测与解决,将MTTR大幅缩短 +- AI驱动根因分析(RCA)可关联跨层问题(计算、网络、应用) +- AI可主动进行预测性维护,从历史宕机模式中学习 +- 智能成本优化可将工作负载动态转移到Spot实例,成本降低40% +- AI驱动的安全审计可自动扫描IAM策略、网络规则、容器漏洞并自动修复 + +## Key Concepts +- [[Agentic AI]]:具有自主决策和任务执行能力的AI系统 +- [[Self-Healing Systems]]:自我修复系统,自动检测并恢复异常 +- [[AIOps]]:AI驱动的运维,利用ML进行异常检测和事件关联 +- [[根因分析]]:识别问题根本原因的系统性方法 + +## Key Entities +- [[Kubernetes]]:容器编排平台(EKS/GKE/AKS) +- [[Terraform]]:IaC工具 +- [[AWS CloudWatch]]:云监控服务 + +## Connections +- [[Agentic AI]] ← 驱动 ← [[AIOps]] +- [[Self-Healing Systems]] ← 依赖 ← 异常检测 +- [[AIOps]] ← 集成 ← 日志分析 + +## Contradictions + diff --git a/wiki/sources/ai-jie-jue-fang-an-zhuan-jia-pei-xun-ke-cheng.md b/wiki/sources/ai-jie-jue-fang-an-zhuan-jia-pei-xun-ke-cheng.md new file mode 100644 index 00000000..da942184 --- /dev/null +++ b/wiki/sources/ai-jie-jue-fang-an-zhuan-jia-pei-xun-ke-cheng.md @@ -0,0 +1,27 @@ +--- +title: "AI 解决方案专家培训课程" +type: source +tags: [coze, AI培训, Agent, 工作流] +date: 2026-04-14 +--- + +## Summary +- 核心主题:Coze平台AI解决方案专家培训课程演示 +- 问题域:如何通过Coze平台构建各行业AI Agent和工作流 +- 方法/机制:提供Coze国内版和海外版多个行业Agent Demo +- 结论/价值:涵盖金融、医疗、教育、电商、客服等多个行业的Coze工作流实践 + +## Key Claims +- Coze平台支持国内版(coze.cn)和海外版(coze.com) +- 行业覆盖:金融(客户分层营销、智能客服)、医疗(分诊助手、影像识别)、教育(知识库问答、拍照搜题)、电商(混剪助手、在线换衣)、人力资源(招聘打分、面试对练) +- AutoGPT和工作流自动化是核心场景 +- 提供多个可复制改造的Agent Demo + +## Key Entities +- [[Coze]]:字节跳动AI Agent平台 + +## Connections +- [[Coze]] ← 培训平台 ← [[AI解决方案专家培训]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/ai-tools-collection.md b/wiki/sources/ai-tools-collection.md new file mode 100644 index 00000000..0add3e2f --- /dev/null +++ b/wiki/sources/ai-tools-collection.md @@ -0,0 +1,28 @@ +--- +title: "我的工具集" +type: source +tags: [AI, Tools, 工具集] +date: 2026-04-14 +source_file: raw/Technical/我的工具集.md +--- + +## Summary +- 核心主题:AI工具清单 +- 类别:Text-to-Speech、Text-to-Video、Image-to-Video、Web-Scraper、AI-Summary + +## Key Tools + +| 类别 | 工具 | 特点 | +|------|------|------| +| Text-to-Speech | Google AI Studio | 免费 | +| Image-to-Video | Wavespeed | 有付费 | +| Image-to-Video | Vidu | $8/月 | +| Image-to-Video | Hailuo | ¥42/月 | +| Web-Scraper | Bright Data | 有付费 | +| AI-Summary | Decopy | 支持文章/PDF/视频摘要 | + +## Connections +- 尚未建立 + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/aionui-cowork-desktop.md b/wiki/sources/aionui-cowork-desktop.md new file mode 100644 index 00000000..7eaf70e2 --- /dev/null +++ b/wiki/sources/aionui-cowork-desktop.md @@ -0,0 +1,32 @@ +--- +title: "OpenClaw as Desktop Cowork (AionUi) — Remote Rescue & Multi-Agent Hub" +type: source +tags: [OpenClaw, AionUi, Desktop Agent, Remote Access, Multi-Agent] +sources: ["https://github.com/iOfficeAI/AionUi"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:OpenClaw 桌面协作与远程救援方案 +- 问题域:如何从桌面 UI 使用 OpenClaw 并远程修复 +- 方法/机制:AionUi 提供多 Agent 协作界面 + 内置 OpenClaw 部署专家 +- 结论/价值:通过 AionUi 在桌面使用 OpenClaw,支持 Telegram/WebUI 远程救援 + +## Key Claims +- AionUi 将 OpenClaw 作为一等桌面 Agent 运行,支持可视化工作区 +- 内置 OpenClaw 部署专家可远程运行 openclaw doctor 修复配置 +- 支持多 Agent 并行:OpenClaw、Claude Code、Codex 等 12+ Agent +- MCP 配置一次,所有 Agent 同步共享 + +## Key Concepts +- [[Desktop Cowork]]:在桌面 UI 中使用 Agent,可视化文件/终端/浏览器操作 +- [[Remote Rescue]]:通过 Telegram/WebUI 远程运行 openclaw doctor 修复 OpenClaw +- [[Multi-Agent Hub]]:AionUi 同时运行 12+ Agent,统一 MCP 配置 + +## Connections +- [[OpenClaw]] ← 核心 ← [[AionUi]] +- [[AionUi]] ← 支持 ← [[Remote Rescue]] +- [[MCP]] ← 共享 ← [[AionUi]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/an-zhuang-ubuntu-24-04-2-zai-hp-zbook-gong-zuo-zhan-bi-ji-ben-shang.md b/wiki/sources/an-zhuang-ubuntu-24-04-2-zai-hp-zbook-gong-zuo-zhan-bi-ji-ben-shang.md new file mode 100644 index 00000000..ba95650e --- /dev/null +++ b/wiki/sources/an-zhuang-ubuntu-24-04-2-zai-hp-zbook-gong-zuo-zhan-bi-ji-ben-shang.md @@ -0,0 +1,33 @@ +--- +title: "安装Ubuntu-24.04.2在HP Zbook工作站笔记本上" +type: source +tags: [Ubuntu, HP, ZBook, Rufus, BIOS] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠安装Ubuntu-24.04.2在HP Zbook工作站笔记本上.md +--- + +## Summary +- 核心主题:在HP ZBook工作站上安装Ubuntu 24.04.2的完整指南 +- 问题域:GPT分区表配置、NVMe优化、BIOS/UEFI设置 +- 结论/价值:详细解决HP ZBook安装Ubuntu后的启动引导问题 + +## Key Claims +- 必须使用GPT分区方案配合UEFI启动 +- BIOS需设置SATA模式为AHCI(而非RAID/Intel RST) +- Secure Boot建议关闭避免第三方驱动问题 +- 解决HP BIOS固执不保存自定义启动项问题 + +## Key Concepts +- [[GPT分区表]]:支持2TB以上硬盘,与UEFI引导完美兼容 +- [[NVMe优化]]:Ubuntu自动识别并对齐优化 +- [[efibootmgr]]:NVRAM强制写入修复启动顺序 + +## Key Entities +- [[HP-ZBook]]:高性能工作站 +- [[Rufus]]:启动盘制作工具 + +## Connections +- [[Ubuntu-Server]] ← 安装 ← [[HP-ZBook-Ubuntu安装]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/an-zhuang-v2rayn.md b/wiki/sources/an-zhuang-v2rayn.md new file mode 100644 index 00000000..46388b81 --- /dev/null +++ b/wiki/sources/an-zhuang-v2rayn.md @@ -0,0 +1,30 @@ +--- +title: "安装v2rayN" +type: source +tags: [v2rayN, Linux, Windows, macOS] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠安装v2rayN.md +--- + +## Summary +- 核心主题:v2rayN跨平台安装指南 +- 问题域:多系统代理客户端配置 +- 结论/价值:支持Windows/macOS/Linux多平台 + +## Key Claims +- v2rayN发布包含Xray、sing-box、mihomo核心 +- Linux支持deb/rpm格式包安装 +- macOS 12+支持,需执行xattr -cr解除损坏提示 + +## Key Concepts +- [[代理核心]]:Xray/sing-box/mihomo +- [[跨平台]]:Windows/macOS/Linux + +## Key Entities +- [[v2rayN]]:代理工具 + +## Connections +- [[Ubuntu-Server]] ← 工具 ← [[v2rayN安装]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/arxiv-paper-reader.md b/wiki/sources/arxiv-paper-reader.md new file mode 100644 index 00000000..54352959 --- /dev/null +++ b/wiki/sources/arxiv-paper-reader.md @@ -0,0 +1,31 @@ +--- +title: "arXiv Paper Reader" +type: source +tags: [arXiv, Research, Paper Reading, LaTeX] +sources: ["https://github.com/Prismer-AI/Prismer/tree/main/skills/arxiv-reader"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:Agent 化 arXiv 论文阅读助手 +- 问题域:如何高效阅读、分析、比较 arXiv 论文 +- 方法/机制:arxiv-reader skill 提供 fetch/sections/abstract 三个工具 +- 结论/价值:在工作区对话式研究论文,无需切换应用 + +## Key Claims +- 按 ID 获取 arXiv 论文,自动展平 LaTeX 公式为可读文本 +- 先浏览论文结构再决定是否通读 +- 多篇摘要对比,快速筛选阅读清单 +- 结果本地缓存,二次访问即时响应 + +## Key Concepts +- [[arXiv Reader]]:arXiv 论文抓取与解析工具 +- [[LaTeX Flattening]]:自动展平 LaTeX 包含文件 +- [[Research Assistant]]:对话式论文研究 Agent + +## Connections +- [[Prismer]] ← 来源 ← [[arXiv Reader]] +- [[OpenClaw]] ← 运行 ← [[arXiv Reader]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/autonomous-game-dev-pipeline.md b/wiki/sources/autonomous-game-dev-pipeline.md new file mode 100644 index 00000000..26d8830a --- /dev/null +++ b/wiki/sources/autonomous-game-dev-pipeline.md @@ -0,0 +1,32 @@ +--- +title: "Autonomous Educational Game Development Pipeline" +type: source +tags: [OpenClaw, Autonomous Agent, Game Development, HTML5, Pipeline] +sources: ["https://github.com/duberblockito/elbebe"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:游戏开发 Agent 全生命周期管理 +- 问题域:如何让单个开发者高效生产 40+ 教育游戏 +- 方法/机制:Agent 自主执行开发队列,Bugs First 策略,7 分钟一个游戏或修复 +- 结论/价值:去中心化开发队列 + 严格设计规则 = 独立开发者高速产出 + +## Key Claims +- "Bugs First" 策略:Agent 必须优先修复 bugs 文件夹中第一个 bug +- Round Robin 策略平衡不同年龄段内容 +- 每 7 分钟完成 1 个新游戏或 bug 修复 +- 自动更新 CHANGELOG.md、master-game-plan.md、games-list.json + +## Key Concepts +- [[Bugs First]]:优先修复 bug 再做新功能的开发策略 +- [[Round Robin]]:轮询策略,均衡分配不同类型任务 +- [[Autonomous Pipeline]]:无人值守自动执行开发队列 +- [[HTML5 Game]]:纯 HTML/CSS/JS 开发,无框架,移动优先 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Autonomous Game Dev]] +- [[Autonomous Project Management]] ← 类似 ← [[Autonomous Game Dev Pipeline]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/autonomous-project-management.md b/wiki/sources/autonomous-project-management.md new file mode 100644 index 00000000..b7ae01fa --- /dev/null +++ b/wiki/sources/autonomous-project-management.md @@ -0,0 +1,33 @@ +--- +title: "Autonomous Project Management with Subagents" +type: source +tags: [OpenClaw, Subagent, Project Management, STATE.yaml, Decentralized] +sources: ["https://www.anthropic.com/research/building-effective-agents"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:去中心化子 Agent 项目管理模式 +- 问题域:如何避免中心化编排瓶颈,实现多任务并行 +- 方法/机制:STATE.yaml 作为共享状态文件,子 Agent 自主更新 +- 结论/价值:CEO 模式(主会话只做策略)+ 子 Agent 自主执行 + +## Key Claims +- 去中心化协调:Agent 读写共享 STATE.yaml 文件 +- 并行执行:多个子 Agent 同时处理独立任务 +- 无编排开销:主会话保持精简(CEO 模式) +- 自我文档化:所有任务状态持久化到版本控制文件 + +## Key Concepts +- [[STATE.yaml]]:项目协调文件,单一真相来源 +- [[CEO Pattern]]:主会话只做策略调度,不做执行 +- [[Subagent]]:子 Agent,自主完成任务并更新状态 +- [[Decentralized Coordination]]:去中心化协调,避免单点瓶颈 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Autonomous PM]] +- [[Autonomous Game Dev Pipeline]] ← 类似 ← [[Autonomous PM]] +- [[Subagent]] ← 实现 ← [[Autonomous PM]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/best-7-news-api-data-feeds.md b/wiki/sources/best-7-news-api-data-feeds.md new file mode 100644 index 00000000..808ec035 --- /dev/null +++ b/wiki/sources/best-7-news-api-data-feeds.md @@ -0,0 +1,58 @@ +--- +title: "Best 7 news API data feeds - AI News" +type: source +tags: [news-api, data-feed, ai] +sources: ["https://www.artificialintelligence-news.com/news/best-7-news-api-data-feeds/"] +last_updated: 2025-03-14 +--- + +## Summary + +- 核心主题:七大新闻API数据源服务对比与选型指南 +- 问题域:企业获取结构化新闻数据的解决方案 +- 方法/机制:API接口聚合多源新闻数据,转化为JSON/XML机器可读格式 +- 结论/价值:帮助企业、开发者根据行业需求选择最适合的新闻API服务 + +## Key Claims + +- 新闻API数据源是聚合、整理并传递来自多个来源的结构化新闻数据的平台 +- API消除了手动收集整理数据的工作,将内容格式化为JSON或XML等机器可读格式 +- 不同API在覆盖范围、价格、专业化程度方面存在显著差异 + +## Key Quotes + +> "Businesses, developers, and analysts rely on news API data feeds to gather structured insights from various sources." — 文章开篇 + +## Key Concepts + +- [[News API]]:新闻API数据源,聚合多源新闻为结构化数据格式 +- [[Webz.io]]:覆盖开放网、深网和暗网的综合性新闻API +- [[GNews API]]:轻量级全球化新闻API,适合小型应用和创业公司 +- [[Guardian API]]:提供高质量编辑内容的新闻API +- [[Bloomberg API]]:专注于金融市场的实时数据API +- [[Financial Times API]]:提供商业和经济深度报道的优质API +- [[Opoint]]:专注于新闻媒体监测和情感分析的API +- [[Mediastack]]:聚合全球7500+来源的可扩展新闻API + +## Key Entities + +- [[Webz.io]]:综合性新闻API提供商 +- [[GNews]]:轻量级新闻API提供商 +- [[The Guardian]]:高质量新闻来源 +- [[Bloomberg]]:金融数据权威提供商 +- [[Financial Times]]:商业与经济新闻来源 + +## Connections + +- [[News API]] ← 应用场景 ← [[金融 intelligence]] +- [[News API]] ← 应用场景 ← [[媒体监测]] +- [[News API]] ← 应用场景 ← [[风险评估]] +- [[News API]] ← 应用场景 ← [[内容聚合平台]] +- [[News API]] ← 应用场景 ← [[AI预测分析]] + +## Contradictions + +- 与其他来源冲突: + - 冲突点:不同API的价格和覆盖范围差异 + - 当前观点:选择API应根据具体行业需求 + - 对方观点:价格是唯一决定因素 diff --git a/wiki/sources/bu-hui-gemini-de-chan-pin-jing-li-zhen-de-yao-bei-tao-tai-le.md b/wiki/sources/bu-hui-gemini-de-chan-pin-jing-li-zhen-de-yao-bei-tao-tai-le.md new file mode 100644 index 00000000..7a718377 --- /dev/null +++ b/wiki/sources/bu-hui-gemini-de-chan-pin-jing-li-zhen-de-yao-bei-tao-tai-le.md @@ -0,0 +1,52 @@ +--- +title: 不会Gemini的产品经理真的要被淘汰了 | 附保姆级PRD生成指南 +type: source +tags: [Gemini, 产品经理, PRD, AI工具] +date: 2025-11-19 +source_file: ../raw/AI/不会Gemini的产品经理真的要被淘汰了 附保姆级PRD生成指南.md +--- + +## Summary +- 核心主题:AI大模型(Gemini)在产品经理日常工作中的应用 +- 问题域:产品经理如何利用大模型提升需求文档编写效率 +- 方法/机制:FeatureList构思 → 逻辑图绘制(ER图、时序图) → PRD撰写 → HTML原型生成 +- 结论/价值:大模型可将产品经理文本工作时间缩短90%以上,但核心"思考"仍需人类完成 + +## Key Claims +- Gemini 2.5可将某些工作时间缩短90%以上 +- 不会用大模型的中初阶产品经理能力结构需重塑 +- 大模型只负责"写"而非"想",人类需完成需求创意的"思考"部分 +- [[FeatureList]]是连接人类思考与大模型输出的有效工具 +- Gemini配合mermaid可生成ER图、时序图、甘特图等多种逻辑图 +- PRD需分页面逐一描述,保持任务难度在Gemini胜任范围内 +- 用"调教"方式训练大模型,三句话可带出一个文档写得好产品经理 + +## Key Quotes +> "只有提交真实需求,才能获得真实的触动" — 纯银 + +> "Gemini是一个知识渊博但'不带脑子'的苦工,你表述的越准、它执行得越准" + +> "三句话,带出来一个文档写得好的产品经理" + +## Key Concepts +- [[FeatureList]]:按层级的需求表,用于构思需求框架 +- [[ER图]]:描述实体、属性、联系的逻辑图,用mermaid语法生成 +- [[时序图]]:表示工作流的逻辑图,需正确使用泳道图语法 +- [[PRD]]:产品需求文档,大模型可辅助撰写但需人类把关 +- [[mermaid]]:图表生成语法,支持ER图、时序图、甘特图等 +- [[Vibe Coding]]:AI编程方式,已被广泛应用 + +## Key Entities +- [[Gemini]]:Google大模型,本文主要使用的AI工具 +- [[Gemini 3 Pro]]:Gemini最新版本 +- [[DeepSider]]:浏览器插件,可访问Nano Banana 2等模型 +- [[飞书]]:本文中用于文档协作和mermaid图表渲染 + +## Connections +- [[Gemini]] ← used_in ← [[PRD生成]] +- [[FeatureList]] ← depends_on ← [[需求构思]] +- [[mermaid]] ← enables ← [[ER图]] +- [[mermaid]] ← enables ← [[时序图]] + +## Contradictions +- 与纯银观点:本文作者对AI商业价值爆发时间线更乐观 diff --git a/wiki/sources/bu-tan-ji-shu.md b/wiki/sources/bu-tan-ji-shu.md new file mode 100644 index 00000000..a6264612 --- /dev/null +++ b/wiki/sources/bu-tan-ji-shu.md @@ -0,0 +1,38 @@ +--- +title: "不谈技术:普通人该怎么在AI时代赚钱?" +type: source +tags: [AI, 职业发展, 思维方式] +date: 2026-03-25 +source_file: raw/微信公众号/不谈技术:普通人该怎么在AI时代赚钱?.md +--- + +## Summary +- 核心主题:AI时代普通人如何找到赚钱路径 +- 问题域:职业发展与个人定位 +- 方法/机制:三大原则(品味值钱、做端到端、用死亡过滤器) +- 结论/价值:AI放大有品味、有方向的人的竞争力 + +## Key Claims +- AI工具民主化,但品味没有民主化——品味是护城河 +- 做端到端的事(从想法到产品),不做别人AI流水线上的零件 +- 用"死亡过滤器"找到真正值得做的事:对一千件事说No,只对一件事说Yes + +## Key Quotes +> "AI不会让普通人变富。AI会让那些知道自己要做什么、并且对品质有执念的人变得极其强大。" — 乔布斯.skill + +> "一个人用AI做出一个完整的App,比一个100人的团队里当'AI提示词工程师'强一万倍。" — 乔布斯.skill + +## Key Concepts +- [[品味值链]]:工具民主化后,品味成为差异化竞争力 +- [[端到端能力]]:从零到一构建完整产品的能力 +- [[死亡过滤器]]:用"如果今天是最后一天"问题筛选真正重要的事 + +## Key Entities +- [[乔布斯.skill]]:提供职业发展建议的AI技能 + +## Connections +- [[养虾日记1]] ← 实证案例 ← [[bu-tan-ji-shu]](AI帮助整理28万张照片是品味+执行力结合的体现) +- [[养虾日记3]] ← 工具支撑 ← [[bu-tan-ji-shu]](Obsidian+Gitea是端到端能力的基础设施) + +## Contradictions +- 与主流"学AI技能"叙事冲突:本文认为零件思维(学工具)不如成品思维(做产品) diff --git a/wiki/sources/build-your-own-x.md b/wiki/sources/build-your-own-x.md new file mode 100644 index 00000000..bc564097 --- /dev/null +++ b/wiki/sources/build-your-own-x.md @@ -0,0 +1,29 @@ +--- +title: "Codecrafters build-your-own-x" +type: source +tags: [build-your-own-x, 学习, 编程, GitHub资源] +sources: ["https://github.com/codecrafters-io/build-your-own-x"] +date: 2026-01-01 +--- + +## Summary +- 核心主题:通过从零复现技术来掌握编程 +- 问题域:如何深入理解现有技术的内部原理 +- 方法/机制:提供25+技术的分步指南,包含3D渲染器、数据库、Docker、Git、操作系统等 +- 结论/价值:Feynman名言"我不能创造的,我也不理解" + +## Key Claims +- 核心理念:What I cannot create, I do not understand +- 涵盖25+技术领域:3D Renderer、Web Server、Docker、Git、Database、Blockchain、Neural Network等 +- 每种技术提供多种编程语言的实现教程 + +## Key Concepts +- [[从零构建]]:通过复现技术来学习的理念 +- [[Build Your Own X]]:涵盖25+技术的实践教程集合 + +## Connections +- [[从零构建]] ← 资源 ← [[Codecrafters]] +- [[Vibe Coding]] ← 相关 ← [[Build Your Own X]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/chatgpt-knowledge-to-canva-gamma.md b/wiki/sources/chatgpt-knowledge-to-canva-gamma.md new file mode 100644 index 00000000..644709dc --- /dev/null +++ b/wiki/sources/chatgpt-knowledge-to-canva-gamma.md @@ -0,0 +1,36 @@ +--- +title: "教學 ChatGPT 先做知識整理,再讓 Canva、 Gamma AI 輸出簡報" +type: source +tags: [AI, ChatGPT, Canva, Gamma, 简报, 知识管理] +date: 2025-10-26 +source_file: raw/Technical/教學 ChatGPT 先做知識整理,再讓 Canva、 Gamma AI 輸出簡報.md +--- + +## Summary +- 核心主题:AI简报制作流程——先ChatGPT知识整理,再Canva/Gamma设计 +- 问题域:直接在Canva/Gamma凭题制作简报容易出现幻觉、内容不够深入 +- 方法/机制:阶段一(5分钟资料研究)→ 阶段二(1分钟建立知识架构)→ 阶段三(1分钟输出简报大纲)→ 阶段四(粘贴到Canva/Gamma) +- 结论/价值:简报不是从版面设计开始,而是从资料研究开始 + +## Key Claims +- 简报制作流程:先在ChatGPT做资料收集、整理、分析,再让Canva/Gamma做出美美版面 +- 阶段一:用5分钟让ChatGPT上网搜索,调阅10笔以上资料作为素材库 +- 阶段二:让ChatGPT建立知识架构,对主题有客观资料认识和主观诠释角度 +- 阶段三:让ChatGPT根据阅读理解输出文字版简报大纲 +- 阶段四:把简报大纲粘贴到Canva/Gamma,利用AI制作简报版面 + +## Key Concepts +- [[知识整理先行]]:简报前期的资料研究、整理、分析是必要步骤 +- [[防弹笔记法]]:任务导向+动态演化+简单精准的笔记系统 +- [[SSOT]]:Single Source of Truth,每个任务一则笔记 + +## Key Entities +- [[ChatGPT]]:用于知识整理和大纲生成 +- [[Canva]]:简报设计工具 +- [[Gamma]]:AI简报工具 + +## Connections +- [[Claude-Prompt库]] ← 知识整理 ← [[ChatGPT-Canva-Gamma简报流程]] + +## Contradictions +- 与"直接用AI生成简报"思维冲突:本文认为前期知识整理不可省略 diff --git a/wiki/sources/china-textbook.md b/wiki/sources/china-textbook.md new file mode 100644 index 00000000..14ef0b49 --- /dev/null +++ b/wiki/sources/china-textbook.md @@ -0,0 +1,27 @@ +--- +title: "ChinaTextbook - 41.53 GB,中国小学、初中、高中、大学 PDF 教材" +type: source +tags: [资源, PDF, 教育, GitHub] +date: 2025-05-13 +source_file: raw/Technical/ChinaTextbook - 41.53 GB,中国小学、初中、高中、大学 PDF 教材.md +--- + +## Summary +- 核心主题:开源项目ChinaTextbook,收集中国K12和大学PDF教材 +- 问题域:教育资源获取 +- 结论/价值:41.53GB免费教材,来源国家中小学智慧教育平台 + +## Key Claims +- 项目地址:github.com/TapXWorld/ChinaTextbook +- 教材来源:国家中小学智慧教育平台 +- 可用第三方工具下载(如tchMaterial-parser) + +## Key Entities +- [[ChinaTextbook]]:开源教材收集项目 +- [[GitHub]]:项目托管平台 + +## Connections +- 尚未建立 + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/claude-code-templates-skills.md b/wiki/sources/claude-code-templates-skills.md new file mode 100644 index 00000000..b1808d94 --- /dev/null +++ b/wiki/sources/claude-code-templates-skills.md @@ -0,0 +1,32 @@ +--- +id: claude-code-templates-skills +title: "在项目里安装 Claude-Code-Templates Skills" +type: source +tags: [claude-code, claude-skills, trae] +sources: [] +last_updated: 2026-04-14 +--- + +## Summary +- 核心主题:Claude Code Templates Skills 的安装和使用方法 +- 问题域:如何快速为项目添加可复用的 Claude Code 技能模板 +- 方法/机制:通过 npx 命令直接安装预设的 Skill 模板 +- 结论/价值:Claude Code Templates 提供开箱即用的 Skill、Agent、MCP 模板 + +## Key Claims +- Claude Code Templates 通过 `npx claude-code-templates@latest --skill=<path> --yes` 命令安装 +- Templates 包含三类资源:Skills(技能)、Agents(代理)、MCP(模型上下文协议) +- Skills 可增强 Claude Code 的特定能力,如 git-commit-helper + +## Key Concepts +- [[Claude-Code-Templates]]:Claude Code 官方模板平台,提供可复用的 Skills、Agents、MCP 资源 +- [[Skill]]:Claude Code 中的可复用技能模块 + +## Key Entities +- [[Claude]]:Anthropic 开发的 AI 助手,Claude Code 的底层模型 + +## Connections +- [[Claude-Code-Templates]] ← 托管于 ← [[aitmpl-com]] +- [[Skill]] ← 组成 ← [[Claude-Code-Templates]] + +## Contradictions diff --git a/wiki/sources/clonezilla-dui-ubuntu-server-jin-hang-quan-pan-jing-xiang-bei-fen.md b/wiki/sources/clonezilla-dui-ubuntu-server-jin-hang-quan-pan-jing-xiang-bei-fen.md new file mode 100644 index 00000000..944c05dc --- /dev/null +++ b/wiki/sources/clonezilla-dui-ubuntu-server-jin-hang-quan-pan-jing-xiang-bei-fen.md @@ -0,0 +1,34 @@ +--- +title: "Clonezilla对Ubuntu Server进行全盘镜像备份" +type: source +tags: [backup, clonezilla, nas, rufus, ubuntu] +date: 2026-04-13 +source_file: raw/Technical/Home Office/🟠Clonezilla对Ubuntu Server进行全盘镜像备份.md +--- + +## Summary +- 核心主题:使用Clonezilla对Ubuntu Server进行全盘镜像备份到NAS +- 问题域:服务器灾难恢复与数据保护 +- 方法/机制:Clonezilla启动盘制作 + NFS网络备份 + ghost风格镜像克隆 +- 结论/价值:提供从启动盘制作到灾难恢复的完整操作流程 + +## Key Claims +- Rufus制作Clonezilla启动盘时必须选择"以ISO镜像模式写入",而非DD镜像模式 +- 针对新笔记本推荐GPT分区方案+UEFI目标系统,旧笔记本使用MBR+BIOS +- NFS网络备份需要正确填写NAS的IP地址和共享路径 +- savedisk模式将整个磁盘备份为镜像文件,压缩选项-z1p提供高压缩率 + +## Key Concepts +- [[磁盘镜像备份]]:将整个磁盘内容复制为镜像文件,用于灾难恢复 +- [[NFS网络存储]]:通过网络文件系统协议挂载远程存储进行备份 + +## Key Entities +- [[Clonezilla]]:开源磁盘克隆工具,类似Ghost的备份恢复软件 +- [[Rufus]]:U盘启动盘制作工具,支持ISO写入 +- [[Ubuntu-Server]]:备份的源系统目标 + +## Connections +- [[Clonezilla]] ← 用于 ← [[磁盘镜像备份]] +- [[磁盘镜像备份]] ← 目标 ← [[Ubuntu-Server]] + +## Contradictions diff --git a/wiki/sources/cloud-computing-myths.md b/wiki/sources/cloud-computing-myths.md new file mode 100644 index 00000000..2cd7bc5b --- /dev/null +++ b/wiki/sources/cloud-computing-myths.md @@ -0,0 +1,34 @@ +--- +title: "The Myths and Misconceptions About Cloud Computing" +type: source +tags: [Cloud Computing, Myths, Security] +sources: [] +last_updated: 2025-03-02 +--- + +## Summary +- 核心主题:澄清关于云计算的常见误解 +- 问题域:消除对云计算安全、成本、控制权的认知偏差 +- 方法/机制:通过事实对比Myth与Reality +- 结论/价值:云计算实际提供增强的安全性、成本效益、可扩展性和数据控制 + +## Key Claims +- Myth 1:云计算不安全 → Reality:云安全通常比本地解决方案更健壮,云提供商投资于加密、防火墙、MFA +- Myth 2:云只是别人的电脑 → Reality:云是具备冗余、可扩展性、高可用性的先进数据中心网络 +- Myth 3:云计算太贵 → Reality:按需付费模型、Reserved Instances、自动扩展和Serverless可显著降低成本 +- Myth 4:失去数据控制权 → Reality:云提供强大的数据治理工具、权限管理、加密和访问监控 +- Myth 5:云只适合大企业 → Reality:中小企业可利用灵活定价获得企业级技术 + +## Key Concepts +- [[云计算]]:通过互联网远程访问计算资源的服务模式 +- [[按需付费]]:根据实际使用量付费的定价模型 +- [[多因素认证]]:Multiple Factor Authentication,增强安全措施 + +## Key Entities + +## Connections +- [[云计算]] ← 包含 ← 安全机制 +- [[按需付费]] ← 实现 ← 成本优化 + +## Contradictions + diff --git a/wiki/sources/cloud-devop-maturity-guideline.md b/wiki/sources/cloud-devop-maturity-guideline.md new file mode 100644 index 00000000..c928f0f3 --- /dev/null +++ b/wiki/sources/cloud-devop-maturity-guideline.md @@ -0,0 +1,37 @@ +--- +title: "Cloud DevOp Maturity - Guideline" +type: source +tags: [DevOps, Cloud, Maturity] +sources: [] +last_updated: 2025-03-01 +--- + +## Summary +- 核心主题:企业级SaaS公司云DevOps成熟度评估框架 +- 问题域:如何评估和提升云DevOps成熟度 +- 方法/机制:基于DORA指标、CMMI模型的成熟度评估体系 +- 结论/价值:提供系统性DevOps成熟度提升路径 + +## Key Claims +- DevOps成熟度定义包含自动化、开发运营协作、交付速度和可靠性四个维度 +- DORA指标包含部署频率、变更前置时间、变更失败率和平均恢复时间(MTTR) +- 成熟度评估应涵盖文化、自动化、结构流程、协作和技术五个关键领域 +- DevOps是持续改进过程,即使成熟公司也需适应新技术和实践 + +## Key Concepts +- [[DevOps成熟度模型]]:评估组织DevOps实践的阶段性框架 +- [[DORA指标]]:DevOps Research & Assessment指标,用于衡量组织DevOps性能 +- [[CMMI]]:Capability Maturity Model Integration,能力成熟度模型集成 +- [[IaC]]:Infrastructure as Code,基础设施即代码 +- [[DevSecOps]]:将安全集成到DevOps生命周期中的实践 + +## Key Entities +- [[DORA]]:DevOps Research & Assessment研究组织,提出四项关键指标 + +## Connections +- [[DevOps成熟度模型]] ← 建立在 ← [[DORA指标]] +- [[DevSecOps]] ← 扩展 ← [[DevOps成熟度模型]] +- [[IaC]] ← 支撑 ← 自动化成熟度 + +## Contradictions + diff --git a/wiki/sources/cloud-maturity-model-guide.md b/wiki/sources/cloud-maturity-model-guide.md new file mode 100644 index 00000000..d2b57763 --- /dev/null +++ b/wiki/sources/cloud-maturity-model-guide.md @@ -0,0 +1,41 @@ +--- +title: "Cloud Maturity Model A Detailed Guide For Cloud Adoption" +type: source +tags: [Cloud, Maturity, Migration] +sources: [] +last_updated: 2025-02-28 +--- + +## Summary +- 核心主题:云成熟度模型(CMM)评估框架与云采用最佳实践 +- 问题域:组织云迁移过程中的成熟度评估与阶段定位 +- 方法/机制:五阶段成熟度模型(0-4级),从完全无云准备到完全优化 +- 结论/价值:帮助组织识别当前云成熟度阶段,制定针对性提升策略 + +## Key Claims +- 云成熟度模型帮助组织从业务和技术两个维度评估云采用准备度 +- 五阶段模型:Level 0(无准备)、Level 1(初始准备)、Level 2(可重复机会主义)、Level 3(系统化文档化)、Level 4(可测量)、Level 5(优化) +- 成熟度提升需平衡人员、流程和技术三个核心维度 +- 云成熟度模型应与业务目标对齐,而非盲目追求最高级别 + +## Key Quotes +> "CMM also helps conduct GAP analysis and identifies areas for improving cloud infrastructure based on business objectives." — OACA定义 + +## Key Concepts +- [[云成熟度模型]]:评估组织云采用能力的阶段性框架 +- [[云迁移策略]]:将工作负载迁移到云环境的方法论 +- [[云治理]]:云环境中的策略、角色和合规性管理 +- [[FinOps]]:云财务运营,用于成本优化 +- [[IaaS/PaaS/SaaS]]:云服务交付模式 + +## Key Entities +- [[OACA]]:Open Alliance for Cloud Adoption,提出云成熟度模型框架 +- [[Gartner]]:研究机构,预测2025年CMM市场规模达15亿美元 + +## Connections +- [[云成熟度模型]] ← 依赖 ← 业务目标对齐 +- [[云治理]] ← 贯穿 ← 所有成熟度级别 +- [[FinOps]] ← 支持 ← Level 4-5成熟度 + +## Contradictions + diff --git a/wiki/sources/cloud-operating-model.md b/wiki/sources/cloud-operating-model.md new file mode 100644 index 00000000..5f9a5cb3 --- /dev/null +++ b/wiki/sources/cloud-operating-model.md @@ -0,0 +1,38 @@ +--- +title: "Cloud Operating Model Key Strategies and Best Practices" +type: source +tags: [Cloud, Operating Model, Governance, DevOps] +sources: [] +last_updated: 2025-03-01 +--- + +## Summary +- 核心主题:云运营模型(COM)设计框架与最佳实践 +- 问题域:组织如何构建结构化云运营体系以实现治理、安全和成本效率 +- 方法/机制:四大核心支柱模型(治理、自动化、安全、成本管理) +- 结论/价值:云运营模型是现代云战略的基础框架,不可或缺 + +## Key Claims +- 89%组织将在2025年采用云优先架构,但仅有结构化方法才能避免成本和安全问题 +- 云运营模型四大支柱:治理与合规、自动化与编排、安全与风险管理、云财务管理(FinOps) +- 59%企业在云成本管理上存在困难,8%关注可持续性和碳足迹 +- 成功云运营需要跨行业定制,金融、医疗、零售、SaaS各有不同需求 + +## Key Concepts +- [[云运营模型]]:管理云资源、安全、自动化和成本的标准化框架 +- [[FinOps]]:云财务运营,平衡云成本与业务价值 +- [[Zero Trust安全模型]]:无隐式信任,持续验证的 security 架构 +- [[IaC]]:Infrastructure as Code,基础设施即代码 +- [[AIOps]]:AI驱动的运维优化 + +## Key Entities +- [[Gartner]]:预测云将成为数字体验的核心 +- [[Flexera]]:发布2024年云状态报告 + +## Connections +- [[云运营模型]] ← 依赖 ← [[FinOps]] +- [[Zero Trust安全模型]] ← 集成 ← [[云运营模型]] +- [[IaC]] ← 支撑 ← 自动化与编排 + +## Contradictions + diff --git a/wiki/sources/cloud-service-delivery.md b/wiki/sources/cloud-service-delivery.md new file mode 100644 index 00000000..714615a5 --- /dev/null +++ b/wiki/sources/cloud-service-delivery.md @@ -0,0 +1,34 @@ +--- +title: "What I know about Cloud Service Delivery" +type: source +tags: [Cloud, Service Delivery, DevOps] +sources: [] +last_updated: 2025-03-01 +--- + +## Summary +- 核心主题:云服务交付的完整生命周期管理 +- 问题域:将云技术能力(IaaS、PaaS、SaaS)转化为可靠、安全、高效的服务 +- 方法/机制:涵盖12个关键领域的云服务交付框架 +- 结论/价值:云服务交付团队需具备多维度技能,包括基础设施、安全、FinOps + +## Key Claims +- 云服务交付团队角色:云基础设施工程师、云运维工程师(DevOps/SRE)、云安全专家、云支持工程师、云FinOps工程师 +- 12个关键领域:服务供应与部署、基础设施管理、平台管理、应用运维、安全与合规管理、性能与可用性监控、事件与问题管理、变更与配置管理、成本管理与优化、客户入职与支持、服务治理与生命周期管理、备份恢复与灾难管理 + +## Key Concepts +- [[云服务交付]]:将云能力转化为最终用户可用服务的完整生命周期 +- [[FinOps]]:云财务管理,优化云成本 +- [[SRE]]:Site Reliability Engineering,网站可靠性工程 +- [[SLA]]:Service Level Agreement,服务级别协议 +- [[SLO]]:Service Level Objective,服务级别目标 + +## Key Entities + +## Connections +- [[云服务交付]] ← 包含 ← [[FinOps]] +- [[SRE]] ← 支撑 ← 运维 +- [[SLA]] ← 定义 ← 服务标准 + +## Contradictions + diff --git a/wiki/sources/content-factory.md b/wiki/sources/content-factory.md new file mode 100644 index 00000000..66c3fe4e --- /dev/null +++ b/wiki/sources/content-factory.md @@ -0,0 +1,32 @@ +--- +title: "Multi-Agent Content Factory" +type: source +tags: [OpenClaw, Multi-Agent, Content Creation, Discord, Pipeline] +sources: ["https://www.youtube.com/watch?v=41_TNGDDnfQ"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:Discord 内多 Agent 内容工厂 +- 问题域:如何自动化完成研究、写作、视觉资产的完整内容流程 +- 方法/机制:Research Agent → Writing Agent → Thumbnail Agent 串联执行 +- 结论/价值:睡前一句话,早上醒来内容已完成 + +## Key Claims +- Research Agent 扫描趋势、竞品、社交媒体寻找内容机会 +- Writing Agent 基于研究结果撰写完整脚本/推文串/简报草稿 +- Thumbnail Agent 生成 AI 缩略图或封面图 +- 每个 Agent 工作在独立 Discord 频道,便于审核反馈 + +## Key Concepts +- [[Content Pipeline]]:研究→写作→视觉资产串联流程 +- [[Multi-Agent Chain]]:多 Agent 接力,一个输出作为下一个输入 +- [[Discord Channel]]:按 Agent 类型划分频道,保持组织清晰 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Content Factory]] +- [[Multi-Agent]] ← 架构 ← [[Content Factory]] +- [[Discord]] ← 平台 ← [[Content Factory]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/cursor-2-0-beginner-guide.md b/wiki/sources/cursor-2-0-beginner-guide.md new file mode 100644 index 00000000..1b58c8cb --- /dev/null +++ b/wiki/sources/cursor-2-0-beginner-guide.md @@ -0,0 +1,34 @@ +--- +title: "Cursor 2.0初学者使用指南" +type: source +tags: [AI, Cursor, IDE, MCP] +date: 2026-04-14 +source_file: raw/Technical/Cursor 2.0初学者使用指南.md +--- + +## Summary +- 核心主题:Cursor 2.0 AI代码编辑器的使用方法 +- 问题域:初学者如何快速上手AI增强代码编辑器 +- 方法/机制:规划 → AI生成计划 → 代码生成 → 代码审查 → 版本控制 +- 结论/价值:Cursor 2.0为开发者提供从想法到实现的智能化路径 + +## Key Claims +- Cursor基于VS Code,免费可用,付费升级获取更多生成额度 +- Composer模型:Cursor自研AI模型,生成速度比同类快4倍 +- 多代理功能可同时运行不同任务,互不干扰 +- Agent模式会修改代码,Ask模式仅提供文本答案 + +## Key Concepts +- [[Composer]]:Cursor自研AI模型 +- [[Agent模式]]:自动执行内嵌命令并处理工具调用 +- [[Ask模式]]:仅提供文本答案,不改动文件 +- [[Diff视图]]:查看AI代码改动对比 + +## Key Entities +- [[Cursor]]:基于VS Code的AI代码编辑器 + +## Connections +- [[MCP在Cursor中的集成与应用]] ← MCP集成 ← [[Cursor-2-0-指南]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/custom-morning-brief.md b/wiki/sources/custom-morning-brief.md new file mode 100644 index 00000000..0fc28240 --- /dev/null +++ b/wiki/sources/custom-morning-brief.md @@ -0,0 +1,32 @@ +--- +title: "Custom Morning Brief" +type: source +tags: [OpenClaw, Morning Brief, Automation, Telegram, Task Management] +sources: ["https://www.youtube.com/watch?v=41_TNGDDnfQ"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:AI 早间简报自动化 +- 问题域:如何利用夜间空闲时间准备次日工作 +- 方法/机制:定时发送个性化报告,整合新闻、任务、AI 推荐任务 +- 结论/价值:将早晨 30 分钟的信息获取时间压缩到起床即读 + +## Key Claims +- 定时发送结构化早间报告(Telegram/Discord/iMessage) +- 研究夜间新闻,聚焦用户兴趣领域 +- 审查任务列表并推送当日待办 +- AI 推荐可自主完成的任务(主动建议而非被动等待指令) + +## Key Concepts +- [[Morning Brief]]:定时自动化早间摘要报告 +- [[Proactive AI]]:主动推荐任务,而非被动响应 +- [[Overnight Processing]]:利用夜间空闲时间生成内容 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Morning Brief]] +- [[Telegram]] ← 通知渠道 ← [[Morning Brief]] +- [[Todoist]] ← 任务集成 ← [[Morning Brief]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/da-mo-xing-xiang-guan-shu-yu-he-kuang-jia-zong-jie.md b/wiki/sources/da-mo-xing-xiang-guan-shu-yu-he-kuang-jia-zong-jie.md new file mode 100644 index 00000000..3ecc13c6 --- /dev/null +++ b/wiki/sources/da-mo-xing-xiang-guan-shu-yu-he-kuang-jia-zong-jie.md @@ -0,0 +1,56 @@ +--- +title: 大模型相关术语和框架总结|LLM、MCP、Prompt、RAG、vLLM、Token、数据蒸馏 +type: source +tags: [LLM, MCP, RAG, vLLM, Token, 数据蒸馏] +date: 2025-12-20 +source_file: ../raw/AI/大模型相关术语和框架总结|LLM、MCP、Prompt、RAG、vLLM、Token、数据蒸馏.md +--- + +## Summary +- 核心主题:大模型核心术语和技术框架 +- 问题域:梳理LLM、MCP、RAG、Agent、LangChain、vLLM、蒸馏等术语 +- 方法/机制:解释各术语的定义、原理和应用场景 +- 结论/价值:为理解大模型技术体系提供基础词汇表 + +## Key Claims +- LLM(Large Language Model):通常以参数规模≥1B被称为"大模型" +- Prompt:输入给大模型的语句 +- [[MCP]](Model Context Protocol):为LLM提供标准化接口连接外部数据源和工具 +- [[Agent]]:大模型+MCP工具融合后的智能体 +- [[RAG]](Retrieval-augmented generation):检索增强生成,解决hallucination问题 +- [[Embedding]]:向量化,将词转化为浮点型数字计算词与词之间的距离 +- [[LangChain]]:快速实现agent的开发框架 +- [[vLLM]]:通过PagedAttention和连续批处理优化GPU内存利用 +- [[KV Cache]]:保存历史K/V向量,避免重复计算 +- Token:大模型基本输入单元,中文约0.6个token/字符 +- 数据蒸馏:用大模型生成精简数据训练小模型 + +## Key Concepts +- [[LLM]]:大语言模型 +- [[MCP]]:模型上下文协议 +- [[Agent]]:智能体 +- [[RAG]]:检索增强生成 +- [[Embedding]]:向量化 +- [[LangChain]]:Agent开发框架 +- [[vLLM]]:高效LLM推理引擎 +- [[KV Cache]]:键值缓存 +- [[Token]]:令牌,大模型基本输入单元 +- [[数据蒸馏]]:知识蒸馏,用大模型数据训练小模型 +- [[PagedAttention]]:vLLM的内存管理技术 +- [[连续批处理]]:vLLM的推理优化技术 +- [[Hallucination]]:大模型一本正经胡说八道的现象 + +## Key Entities +- [[LangChain]]:Agent开发框架 +- [[vLLM]]:开源LLM推理优化项目 + +## Connections +- [[LLM]] ← uses ← [[Token]] +- [[LLM]] ← enhanced_by ← [[RAG]] +- [[LLM]] ← enhanced_by ← [[Agent]] +- [[Agent]] ← built_with ← [[MCP]] +- [[vLLM]] ← uses ← [[PagedAttention]] +- [[vLLM]] ← uses ← [[KV Cache]] +- [[数据蒸馏]] ← generates ← [[LLM]] + +## Contradictions diff --git a/wiki/sources/daily-reddit-digest.md b/wiki/sources/daily-reddit-digest.md new file mode 100644 index 00000000..d5976780 --- /dev/null +++ b/wiki/sources/daily-reddit-digest.md @@ -0,0 +1,30 @@ +--- +title: "Daily Reddit Digest" +type: source +tags: [OpenClaw, Reddit, Digest, Social Media, Automation] +sources: ["https://clawhub.ai/buksan1950/reddit-readonly"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:每日 Reddit 热门内容摘要 +- 问题域:如何高效跟踪多个 subreddit 的热门帖子 +- 方法/机制:reddit-readonly skill 抓取热帖,支持搜索和评论线程 +- 结论/价值:每日定时推送,只读模式不参与互动 + +## Key Claims +- 浏览 subreddit 热门/最新/最高赞帖子 +- 按主题搜索帖子 +- 拉取评论线程获取上下文 +- 构建待手动审核/回复的帖子候选列表 + +## Key Concepts +- [[Reddit Digest]]:Reddit 热门帖子每日摘要 +- [[Read-only Mode]]:只读,不发帖、点赞或评论 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Reddit Digest]] +- [[reddit-readonly skill]] ← 工具 ← [[Reddit Digest]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/daily-youtube-digest.md b/wiki/sources/daily-youtube-digest.md new file mode 100644 index 00000000..4452a461 --- /dev/null +++ b/wiki/sources/daily-youtube-digest.md @@ -0,0 +1,32 @@ +--- +title: "Daily YouTube Digest" +type: source +tags: [OpenClaw, YouTube, Digest, Transcript, Automation] +sources: ["https://clawhub.ai/therohitdas/youtube-full"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:每日 YouTube 视频摘要 +- 问题域:如何不再错过订阅频道的新视频 +- 方法/机制:youtube-full skill 获取最新视频并提取字幕生成摘要 +- 结论/价值:每天定时推送精选频道的新视频摘要 + +## Key Claims +- 获取最喜欢的 YouTube 频道最新视频 +- 提取视频字幕生成关键见解摘要 +- 支持按频道或按关键词两种模式 +- channel/latest 和 channel/resolve 免费,字幕每个 1 credit + +## Key Concepts +- [[YouTube Digest]]:YouTube 频道新视频每日摘要 +- [[Transcript Extraction]]:从视频提取字幕并摘要 +- [[Channel Tracking]]:跟踪指定频道的新上传 + +## Connections +- [[OpenClaw]] ← 运行 ← [[YouTube Digest]] +- [[youtube-full skill]] ← 工具 ← [[YouTube Digest]] +- [[TranscriptAPI]] ← 依赖 ← [[YouTube Digest]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/deepseek-handbook.md b/wiki/sources/deepseek-handbook.md new file mode 100644 index 00000000..aad0cb2f --- /dev/null +++ b/wiki/sources/deepseek-handbook.md @@ -0,0 +1,33 @@ +--- +title: "清华出的DeepSeek使用手册" +type: source +tags: [DeepSeek, AI, 提示词, 教程] +date: 2025-12-18 +source_file: raw/Technical/清华出的DeepSeek使用手册,104页,真的是太厉害了!(免费领取).md +--- + +## Summary +- 核心主题:清华大学DeepSeek使用手册内容介绍 +- 问题域:AI使用效率提升 +- 结论/价值:授人以鱼不如授人以渔,教底层逻辑而非表面操作 + +## Key Claims +- 文档由清华大学余梦珑博士后团队撰写 +- 核心内容:DeepSeek技术特点、应用场景、使用方法、提示语设计 +- 特色:不仅告诉怎么提问,还告诉为什么这么问 +- 价值:104页干货,避免AI幻觉,设计超棒提示语的秘籍 + +## Key Concepts +- [[提示词底层逻辑]]:知道为什么问,才知道怎么问 +- [[AI幻觉避免]]:手册提供的实用技巧 + +## Key Entities +- [[DeepSeek]]:通用人工智能公司 +- [[DeepSeek-R1]]:开源推理模型 +- [[清华大学]]:手册发布机构 + +## Connections +- [[Claude-Prompt库]] ← 提示词工程 ← [[DeepSeek使用手册]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/designing-for-agentic-ai.md b/wiki/sources/designing-for-agentic-ai.md new file mode 100644 index 00000000..016bde32 --- /dev/null +++ b/wiki/sources/designing-for-agentic-ai.md @@ -0,0 +1,57 @@ +--- +title: "Designing for Agentic AI" +type: source +tags: [agentic-ai, product-design, genai] +sources: ["https://www.linkedin.com/pulse/designing-agentic-ai-yuri-pessa-ztcmf/"] +last_updated: 2025-03-02 +--- + +## Summary + +- 核心主题:Agentic AI(智能体AI)产品设计原则与实践 +- 问题域:如何设计能够主动行动、决策的AI系统用户体验 +- 方法/机制:通过透明度、控制权、个性化、对话和预期五大设计原则指导Agentic AI体验设计 +- 结论/价值:为产品设计师提供设计Agentic AI系统的最佳实践框架 + +## Key Claims + +- GenAI擅长生成新内容,Agentic AI专注于行动、决策和预见用户需求 +- Agentic AI引入新维度:主动出击的智能体预见需求并自主行动 +- 用户观察AI决策过程本身也是一种互动形式 +- 设计隐喻需要从被动响应用户动作转向提供实时反馈的体验 + +## Key Quotes + +> "Agentic AI is all about action. It can interact with its environment, make decisions, and even anticipate user needs." — GenAI vs Agentic AI对比 + +> "Users should always feel in control of the AI." — 设计原则 + +## Key Concepts + +- [[Agentic AI]](智能体AI):能够互动、决策并预见用户需求的主动行动型AI +- [[GenAI]](生成式AI):擅长创建新内容的创意助手型AI +- [[透明度原则]]:用户应能理解AI如何做决策 +- [[控制权原则]]:用户应能停止AI行为或撤销AI已执行的动作 +- [[个性化原则]]:Agentic AI应适应个体用户需求和偏好 +- [[对话原则]]:设计自然直观的用户与AI对话界面 +- [[预期原则]]:Agentic AI应能预见用户需求并主动提供帮助 + +## Key Entities + +- [[Yuri Pessa]]:文章作者,AI产品设计师 + +## Connections + +- [[Agentic AI]] ← 对比 ← [[GenAI]] +- [[透明度原则]] ← 属于 ← [[Agentic AI设计原则]] +- [[控制权原则]] ← 属于 ← [[Agentic AI设计原则]] +- [[个性化原则]] ← 属于 ← [[Agentic AI设计原则]] +- [[对话原则]] ← 属于 ← [[Agentic AI设计原则]] +- [[预期原则]] ← 属于 ← [[Agentic AI设计原则]] + +## Contradictions + +- 与传统交互设计冲突: + - 冲突点:用户是否被动 + - 当前观点:用户通过观察AI决策过程保持互动 + - 对方观点:用户完全被动接受AI服务 diff --git a/wiki/sources/dev-experience-standards.md b/wiki/sources/dev-experience-standards.md new file mode 100644 index 00000000..3983a0f4 --- /dev/null +++ b/wiki/sources/dev-experience-standards.md @@ -0,0 +1,38 @@ +--- +title: "开发经验与项目规范整理文档" +type: source +tags: [开发, 规范, 编码, 微服务] +date: 2026-04-14 +source_file: raw/Technical/开发经验与项目规范整理文档.md +--- + +## Summary +- 核心主题:软件开发规范与最佳实践 +- 问题域:变量命名、文件结构、编码规范、系统架构 +- 方法/机制:建立统一索引 → 明确输入输出 → 遵循单一职责 +- 结论/价值:工程化思维减少复杂度、提高可维护性 + +## Key Claims +- 变量名大全文件:统一变量命名,方便全局搜索和AI管理 +- 文件结构:每个子目录包含agents和claude.md说明文档 +- 编码规范:消费端/生产端/状态/变换明确划分 +- 系统架构原则:先梳理清架构,再写代码 +- 程序设计核心:需求 → 简单架构 → 可维护测试 → 小步迭代 +- DRY原则:提炼公共逻辑,避免重复代码 + +## Key Concepts +- [[变量命名规范]]:小写英文+下划线或小驼峰,语义化 +- [[单一职责]]:每个文件/类/函数只负责一件事 +- [[输入输出划分]]:消费端/生产端/状态/变换 +- [[DRY原则]]:Don't Repeat Yourself,不重复代码 +- [[微服务架构]]:独立开发、独立部署、独立扩容 + +## Key Entities +- [[Redis]]:缓存、提升读性能、降低数据库压力 +- [[消息队列]]:异步通信、解耦、削峰填谷 + +## Connections +- [[Trae远程开发部署指南]] ← 开发实践 ← [[开发经验与规范]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/devops-culture-transformation.md b/wiki/sources/devops-culture-transformation.md new file mode 100644 index 00000000..38b56b16 --- /dev/null +++ b/wiki/sources/devops-culture-transformation.md @@ -0,0 +1,38 @@ +--- +title: "DevOps Culture and Transformation: Fostering Collaboration, Agile Practices, and Innovation" +type: source +tags: [DevOps, Culture, Transformation, Agile] +sources: [] +last_updated: 2025-03-02 +--- + +## Summary +- 核心主题:DevOps文化转型与协作机制建设 +- 问题域:如何通过文化转型实现DevOps实践的全面落地 +- 方法/机制:四大DevOps文化支柱(协作、自动化、持续改进、客户中心) +- 结论/价值:DevOps不仅是工具和自动化,更是一种优先考虑协作、持续学习和客户中心的思维方式转变 + +## Key Claims +- DevOps四大基础原则:协作优于孤岛、自动化作为使能器、持续改进(Kaizen)、客户中心 +- 跨功能团队是打破开发者与运维之间壁垒的关键 +- CI/CD是Agile加速器,将反馈周期从数周缩短到数分钟 +- DevOps未来趋势:AI/ML in DevOps、GitOps、无服务器DevOps、边缘计算与IoT DevOps、增强安全DevSecOps + +## Key Concepts +- [[DevOps文化]]:强调协作、自动化和持续改进的组织文化 +- [[CI/CD流水线]]:持续集成/持续交付的自动化构建测试部署流程 +- [[IaC]]:Infrastructure as Code,基础设施即代码 +- [[Kaizen]]:持续改进哲学 +- [[DevSecOps]]:将安全集成到DevOps + +## Key Entities +- [[Atlassian Jira]]:团队协作与工作流管理工具 +- [[Slack]]:团队实时通信平台 + +## Connections +- [[DevOps文化]] ← 依赖 ← 跨功能团队 +- [[CI/CD流水线]] ← 实现 ← 自动化 +- [[DevSecOps]] ← 扩展 ← DevOps文化 + +## Contradictions + diff --git a/wiki/sources/devops-maturity-model.md b/wiki/sources/devops-maturity-model.md new file mode 100644 index 00000000..17618061 --- /dev/null +++ b/wiki/sources/devops-maturity-model.md @@ -0,0 +1,40 @@ +--- +title: "DevOps Maturity Model From Traditional IT to Advanced DevOps" +type: source +tags: [DevOps, Maturity, CI/CD] +sources: [] +last_updated: 2025-03-01 +--- + +## Summary +- 核心主题:DevOps成熟度模型的五阶段演进路径 +- 问题域:组织如何从传统IT逐步演进到高级DevOps实践 +- 方法/机制:五阶段成熟度模型(初始/临时阶段、局部DevOps、自动化定义、高度优化、完全成熟) +- 结论/价值:DevOps成熟度模型是引导组织DevOps转型的重要工具 + +## Key Claims +- 五阶段DevOps成熟度:Phase 1(初始/Ad-Hoc)、Phase 2(局部DevOps)、Phase 3(自动化和定义)、Phase 4(高度优化)、Phase 5(完全成熟) +- Phase 1特征:团队孤立工作、瀑布式方法、手动测试、安全仅在发布前几周介入 +- Phase 5特征:自给自足全栈团队、每日多次部署、零人工干预、安全防止不合规代码进入生产 +- DevOps成熟度关键指标:部署频率、MTTR、变更失败率、代码质量 + +## Key Quotes +> "The DevOps Maturity Model assessment allows organizations to: Analyze and measure their current DevOps capabilities and methodologies." — Bacancy Technology + +## Key Concepts +- [[DevOps成熟度模型]]:评估组织DevOps实践的阶段性框架 +- [[CI/CD]]:持续集成/持续交付 +- [[MTTR]]:Mean Time to Recovery,平均恢复时间 +- [[变更失败率]]:导致生产失败或需要回滚的部署百分比 +- [[技术债务]]:为了加速交付而接受的非最优代码或架构权衡 + +## Key Entities +- [[Bacancy Technology]]:DevOps咨询公司 + +## Connections +- [[DevOps成熟度模型]] ← 包含 ← [[CI/CD]] +- [[MTTR]] ← 衡量 ← 运维绩效 +- [[变更失败率]] ← 衡量 ← 部署质量 + +## Contradictions + diff --git a/wiki/sources/devsecops-best-practices.md b/wiki/sources/devsecops-best-practices.md new file mode 100644 index 00000000..5a60c4c8 --- /dev/null +++ b/wiki/sources/devsecops-best-practices.md @@ -0,0 +1,44 @@ +--- +title: "What is DevSecOps Best Practices, Benefits, and Tools" +type: source +tags: [DevSecOps, Security, SDLC, CI/CD] +sources: [] +last_updated: 2025-12-19 +--- + +## Summary +- 核心主题:DevSecOps实践、收益和工具全面解析 +- 问题域:如何将安全集成到DevOps生命周期的每个阶段 +- 方法/机制:Shift Left安全实践、自动化安全测试、协作文化 +- 结论/价值:70%发布后发现的安全漏洞可通过DevSecOps预防 + +## Key Claims +- DevSecOps代表Development+Security+Operations,将安全集成到整个软件开发生命周期 +- SDLC包含:需求分析、规划、架构设计、开发、测试、部署 +- DevSecOps核心价值:快速发布成本效益、提升主动安全、快速漏洞修复、与现代开发兼容的自动化 +- Shift Left:将安全缺陷识别提前到软件开发生命周期早期 +- 五大组件:协作、沟通、自动化、工具与架构安全、测试 +- 安全测试类型:SAST(静态)、SCA(软件组成分析)、IAST(交互式)、DAST(动态) + +## Key Quotes +> "70% of software vulnerabilities discovered post-launch could have been prevented with DevSecOps" — Bacancy Technology + +## Key Concepts +- [[DevSecOps]]:将安全集成到DevOps的实践 +- [[SDLC]]:Software Development Lifecycle,软件开发生命周期 +- [[Shift Left]]:将安全测试提前到开发早期阶段的实践 +- [[SAST]]:Static Application Security Testing,静态应用安全测试 +- [[DAST]]:Dynamic Application Security Testing,动态应用安全测试 +- [[SCA]]:Software Composition Analysis,软件组成分析 + +## Key Entities +- [[Bacancy Technology]]:DevSecOps咨询公司 + +## Connections +- [[DevSecOps]] ← 包含 ← [[Shift Left]] +- [[SAST]] ← 集成 ← CI/CD +- [[DAST]] ← 集成 ← CI/CD +- [[SDLC]] ← 覆盖 ← DevSecOps + +## Contradictions + diff --git a/wiki/sources/dynamic-dashboard.md b/wiki/sources/dynamic-dashboard.md new file mode 100644 index 00000000..e84301b5 --- /dev/null +++ b/wiki/sources/dynamic-dashboard.md @@ -0,0 +1,30 @@ +--- +title: "Dynamic Dashboard with Sub-agent Spawning" +type: source +tags: [OpenClaw, Subagent, Dashboard, Real-time, Metrics] +date: 2026-04-14 +--- + +## Summary +- 核心主题:动态实时监控面板 +- 问题域:如何并行获取多数据源并统一展示 +- 方法/机制:子 Agent 并行抓取数据,聚合到 Discord 或 HTML 面板 +- 结论/价值:对话式定义监控指标,自动定时更新 + +## Key Claims +- 并行监控多个数据源(API、数据库、GitHub、社交媒体) +- 子 Agent 分散 API 调用避免阻塞和限流 +- 聚合结果统一展示,支持阈值告警 +- 历史趋势存入数据库支持可视化 + +## Key Concepts +- [[Dynamic Dashboard]]:实时多数据源聚合面板 +- [[Sub-agent Spawning]]:并行生成子 Agent 分散工作负载 +- [[Metrics Aggregation]]:多源指标聚合 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Dynamic Dashboard]] +- [[Subagent]] ← 实现 ← [[Dynamic Dashboard]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/earnings-tracker.md b/wiki/sources/earnings-tracker.md new file mode 100644 index 00000000..bb359c20 --- /dev/null +++ b/wiki/sources/earnings-tracker.md @@ -0,0 +1,28 @@ +--- +title: "AI-Powered Earnings Tracker" +type: source +tags: [OpenClaw, Earnings, Finance, Telegram, Cron] +date: 2026-04-14 +--- + +## Summary +- 核心主题:AI 驱动财报追踪 +- 问题域:如何自动跟踪科技公司财报发布和结果 +- 方法/机制:周日预览下周财报日历,指定日期自动抓取并摘要 +- 结论/价值:不错过任何关注的财报 + +## Key Claims +- 周日扫描下周财报日历,过滤关注的科技/AI 公司 +- 用户确认后为每场财报设置一次性 cron 任务 +- 财报发布后自动搜索结果,格式化摘要(beat/miss、关键指标、AI 亮点) + +## Key Concepts +- [[Earnings Tracker]]:财报发布日期和结果追踪 +- [[Cron Job]]:一次性定时任务触发财报摘要 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Earnings Tracker]] +- [[Telegram]] ← 通知渠道 ← [[Earnings Tracker]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/er-chuang-shi-pin-bi-bu-ke-shao-ai-gong-ju-tui-jian-he-ji-ai-pei-yin-sheng-yin-ke-long.md b/wiki/sources/er-chuang-shi-pin-bi-bu-ke-shao-ai-gong-ju-tui-jian-he-ji-ai-pei-yin-sheng-yin-ke-long.md new file mode 100644 index 00000000..b2d117a4 --- /dev/null +++ b/wiki/sources/er-chuang-shi-pin-bi-bu-ke-shao-ai-gong-ju-tui-jian-he-ji-ai-pei-yin-sheng-yin-ke-long.md @@ -0,0 +1,42 @@ +--- +title: 二创视频必不可少!2025年最热门AI工具推荐合集-AI配音、声音克隆 +type: source +tags: [AI配音, 声音克隆, AI工具] +date: 2025-03-06 +source_file: ../raw/AI/二创视频必不可少!2025年最热门AI工具推荐合集-AI配音、声音克隆.md +--- + +## Summary +- 核心主题:AI配音和声音克隆工具推荐 +- 问题域:二创视频制作中如何选择合适的AI配音工具 +- 方法/机制:对比ElevenLabs、海螺AI、F5-TTS、TTSMaker、剪映、魔音工坊、AnyVoice等7款工具 +- 结论/价值:按需求(高品质/免费/技术流/短视频新手)推荐不同工具 + +## Key Claims +- ElevenLabs:国际顶流,支持30+语言和方言,声音自然度高,API接口灵活 +- 海螺AI(MiniMax出品):小白友好,30秒克隆声音,支持中文/粤语等17种语言,免费 +- F5-TTS:程序员专属,开源免费,2秒音频克隆,支持本地部署 +- TTSMaker:打工人必备,每周免费3万字,50+语言、300+音色,生成音频可商用 +- 剪映:抖音官方,视频剪辑无缝衔接,有"小帅""小美"等网红音色 +- 魔音工坊:土豪团队首选,500+音色可选,普通克隆免费 +- AnyVoice:3秒克隆黑科技,免费无限下载,支持中英日韩四语 + +## Key Entities +- [[ElevenLabs]]:国际顶流AI配音平台 +- [[海螺AI]]:MiniMax出品的中文AI配音工具 +- [[F5-TTS]]:开源语音合成平台 +- [[TTSMaker]]:马克配音工具 +- [[剪映]]:字节跳动旗下视频剪辑工具 +- [[魔音工坊]]:声音克隆工具 +- [[AnyVoice]]:多语言声音克隆平台 +- [[MiniMax]]:AI工具出品公司 + +## Key Concepts +- [[声音克隆]]:用少量音频样本复制特定声音的技术 +- [[AI配音]]:使用AI生成语音的技术 + +## Connections +- [[声音克隆]] ← enables ← [[AI配音]] +- [[海螺AI]] ← variant_of ← [[MiniMax]] + +## Contradictions diff --git a/wiki/sources/event-guest-confirmation.md b/wiki/sources/event-guest-confirmation.md new file mode 100644 index 00000000..51f7551d --- /dev/null +++ b/wiki/sources/event-guest-confirmation.md @@ -0,0 +1,32 @@ +--- +title: "Event Guest Confirmation" +type: source +tags: [OpenClaw, SuperCall, Voice AI, Event, Phone Call] +sources: ["https://clawhub.ai/xonder/supercall"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:AI 电话确认活动出席 +- 问题域:如何批量电话确认 20+ 宾客出席信息 +- 方法/机制:SuperCall 逐个拨打电话,收集饮食禁忌、加一等信息,汇总报告 +- 结论/价值:真电话接通率远高于短信 + +## Key Claims +- 逐个拨打电话,AI 自称为活动协调员 +- 确认活动日期、时间、地点,收集出席意愿和备注 +- 完成后汇总:已确认、已拒绝、未接听、备注 +- SuperCall 是独立语音 Agent,无网关访问权限,安全性高 + +## Key Concepts +- [[SuperCall]]:独立语音 Agent,与主网关隔离 +- [[Voice AI]]:AI 电话外呼确认 +- [[Guest Confirmation]]:活动宾客出席确认 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Event Guest Confirmation]] +- [[SuperCall]] ← 工具 ← [[Event Guest Confirmation]] +- [[Twilio]] ← 电话服务 ← [[Event Guest Confirmation]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/family-calendar-household-assistant.md b/wiki/sources/family-calendar-household-assistant.md new file mode 100644 index 00000000..7b74587a --- /dev/null +++ b/wiki/sources/family-calendar-household-assistant.md @@ -0,0 +1,32 @@ +--- +title: "Family Calendar Aggregation & Household Assistant" +type: source +tags: [OpenClaw, Calendar, Household, iMessage, Telegram] +sources: ["https://news.ycombinator.com/item?id=46872465"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:家庭日历聚合与家务助手 +- 问题域:如何整合多个平台的日历和家庭协调信息 +- 方法/机制:AI 被动监控消息创建日历事件,管理库存和购物清单 +- 结论/价值:Mac Mini 运行 iMessage 集成是最佳方案 + +## Key Claims +- 早晨简报聚合所有家庭日历 +- 被动监控 iMessage,自动从文本检测约会并创建日历事件(含行车缓冲时间) +- 家庭库存管理(照片/文本/收据更新) +- 购物清单去重合并 + +## Key Concepts +- [[Calendar Aggregation]]:多日历源整合 +- [[Ambient Monitoring]]:被动监控消息并主动创建事件 +- [[Household Inventory]]:家庭物品库存追踪 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Family Calendar Assistant]] +- [[iMessage]] ← 消息源 ← [[Family Calendar Assistant]] +- [[Google Calendar]] ← 日历源 ← [[Family Calendar Assistant]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/formalization-recursive-self-optimizing-generative-systems.md b/wiki/sources/formalization-recursive-self-optimizing-generative-systems.md new file mode 100644 index 00000000..933073a7 --- /dev/null +++ b/wiki/sources/formalization-recursive-self-optimizing-generative-systems.md @@ -0,0 +1,40 @@ +--- +title: "A Formalization of Recursive Self-Optimizing Generative Systems" +type: source +tags: [递归自优化, 生成系统, 固定点理论, lambda演算, AI理论] +sources: ["https://github.com/2025Emma/vibe-coding-cn/blob/main/i18n/zh/documents/Methodology%20and%20Principles/A%20Formalization%20of%20Recursive%20Self-Optimizing%20Generative%20Systems.md"] +date: 2025-12-30 +--- + +## Summary +- 核心主题:递归自优化生成系统的形式化理论 +- 问题域:如何形式化描述AI系统的自我完善机制 +- 方法/机制:通过固定点理论和λ演算建立递归自优化生成系统的数学模型 +- 结论/价值:证明递归自优化系统自然引导到固定点结构,而非终端输出 + +## Key Claims +- 系统目标不是直接产生最优输出,而是通过迭代自我修改构建稳定的生成能力 +- Generator(α)生成artifacts,Optimizer(Ω)改进artifacts,Meta-generator用优化结果更新Generator +- 稳定生成能力定义为Φ的固定点:G* ∈ G,Φ(G*) = G* +- 递归结构可用λ演算表达:G* = Y·STEP(Y为固定点组合子) +- 系统对齐经典自引用、递归和自举计算结果 + +## Key Concepts +- [[递归自优化]]:通过迭代自我修改构建稳定生成能力 +- [[固定点理论]]:稳定生成能力对应Φ的固定点 +- [[λ演算]]:表达递归自引用动力学 +- [[自举]](Bootstrapping):用自身输出更新自身 +- [[Generator]](α-提示词):生成其他提示词的"母体"提示词 +- [[Optimizer]](Ω-提示词):优化其他提示词的"母体"提示词 + +## Connections +- [[递归自优化]] ← 包含 ← [[Generator]] +- [[递归自优化]] ← 包含 ← [[Optimizer]] +- [[递归自优化]] ← 形式化 ← [[固定点理论]] +- [[递归自优化]] ← 表达 ← [[λ演算]] + +## Contradictions +- 与直接输出优化观点冲突: + - 冲突点:系统目标 + - 当前观点:目标是构建稳定的生成能力,而非最优输出 + - 对方观点:直接优化输出即可 diff --git a/wiki/sources/git-push-connection-reset.md b/wiki/sources/git-push-connection-reset.md new file mode 100644 index 00000000..3a6c6d63 --- /dev/null +++ b/wiki/sources/git-push-connection-reset.md @@ -0,0 +1,31 @@ +--- +title: "Git Push 连接重置问题修复" +type: source +tags: [Git, GitHub, Proxy, SOCKS5] +date: 2026-04-14 +source_file: raw/Technical/Git Push 连接重置问题修复.md +--- + +## Summary +- 核心主题:解决国内访问GitHub时Git Push连接重置问题 +- 问题域:TCP连接层面被防火墙阻断 +- 方法/机制:配置Git代理 / 切换到SSH协议 +- 结论/价值:让Git流量走本地代理通道是根本解决方案 + +## Key Claims +- `Recv failure: Connection was reset`是TCP连接层面中断,不是权限问题 +- GFW检测到流量特征后发送TCP RST包阻断连接 +- 为Git单独配置代理不影响其他命令 + +## Key Concepts +- [[Git代理配置]]:为Git单独设置HTTP/SOCKS5代理 +- [[SSH协议切换]]:从HTTPS切换到SSH连接GitHub + +## Key Entities +- [[GitHub]]:代码托管平台 + +## Connections +- [[Trae远程开发部署指南]] ← 开发协作 ← [[Git-Push-连接重置]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/gu-ding-jing-tou-duan-shi-pin-zhi-zuo-de-ai-quan-liu-cheng-jie-xi.md b/wiki/sources/gu-ding-jing-tou-duan-shi-pin-zhi-zuo-de-ai-quan-liu-cheng-jie-xi.md new file mode 100644 index 00000000..27a850fa --- /dev/null +++ b/wiki/sources/gu-ding-jing-tou-duan-shi-pin-zhi-zuo-de-ai-quan-liu-cheng-jie-xi.md @@ -0,0 +1,46 @@ +--- +title: 固定镜头短视频制作的AI全流程解析 +type: source +tags: [AI视频, 短视频制作, 家装视频] +date: 2026-03-15 +source_file: ../raw/AI/固定镜头短视频制作的AI全流程解析.md +--- + +## Summary +- 核心主题:利用AI技术快速制作高播放量的固定镜头短视频 +- 问题域:如何用AI复刻家装类"从毛坯到精装"效果的视频 +- 方法/机制:文案分镜 → 九宫格图片生成 → 首尾针动画 → 快节奏剪辑 → 声音设计 +- 结论/价值:用AI不到10分钟即可完成成片,核心在于固定机位、内容连续变化、时间压缩 + +## Key Claims +- 家装短视频三大关键词:固定机位、内容连续变化、时间压缩 +- AI工具分类:大脑类(分镜转换)、设计师类(图像生成)、动效类(动画生成) +- 九宫格法可保证画面一致性,同时生成3x3九个分镜画面 +- 首尾针动画:通过首针图和尾针图补齐中间变化 +- 快节奏剪辑:建议2-4倍速加速,采用硬切而非复杂转场 +- 声音设计:施工音效+节奏感强的BGM,画面变化处精准卡点 + +## Key Concepts +- [[固定机位]]:摄像机位置固定不变,是视频画面统一和连贯的基础 +- [[内容连续变化]]:视频主体信息随时间持续发生明确阶段性变化 +- [[时间压缩]]:将长时间拍摄过程在视频中浓缩表现的手法 +- [[分镜拆解]]:将视频内容拆分成多个画面阶段描述 +- [[九宫格法]]:同时生成3x3共九个画面,保证机位与角度不变 +- [[首尾针动画]]:通过上传两个关键帧(首针和尾针),AI自动补齐中间动作 +- [[快节奏剪辑]]:使用加速播放和硬切换手法,强化节奏感 +- [[卡点]]:画面变化与音乐节奏巧妙同步 + +## Key Entities +- [[海螺AI]]:动效类工具 +- [[KAI]]:AI视频生成工具,支持首尾针动画 +- [[Midjourney]]:设计师类图像生成工具 +- [[Nano Banana]]:设计师类图像生成工具 +- [[Google AI Studio]]:大脑类工具,用于分镜拆解 + +## Connections +- [[固定机位]] ← enables ← [[九宫格法]] +- [[九宫格法]] ← generates ← [[首尾针动画]] +- [[首尾针动画]] ← combines_with ← [[快节奏剪辑]] +- [[快节奏剪辑]] ← enhanced_by ← [[卡点]] + +## Contradictions diff --git a/wiki/sources/habit-tracker-accountability-coach.md b/wiki/sources/habit-tracker-accountability-coach.md new file mode 100644 index 00000000..1999b075 --- /dev/null +++ b/wiki/sources/habit-tracker-accountability-coach.md @@ -0,0 +1,30 @@ +--- +title: "Habit Tracker & Accountability Coach" +type: source +tags: [OpenClaw, Habit, Accountability, Telegram, Streak] +date: 2026-04-14 +--- + +## Summary +- 核心主题:主动习惯追踪与问责伙伴 +- 问题域:如何让习惯追踪从被动记录变为主动督促 +- 方法/机制:定时检查询问完成情况,根据连续天数调整语气,生成周报分析模式 +- 结论/价值:AI 主动督促比 App 推送有效 + +## Key Claims +- 每日定时 Telegram/SMS 检查询问习惯完成情况 +- 连续天数追踪,消息中引用当前连续天数 +- 自适应语气:稳定时鼓励,落伍时温和坚持 +- 周报分析完成率、最长连续、发现模式 + +## Key Concepts +- [[Habit Tracker]]:习惯追踪 +- [[Accountability Partner]]:主动督促伙伴 +- [[Adaptive Nudge]]:基于表现调整的提醒语气 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Habit Tracker]] +- [[Telegram]] ← 通知渠道 ← [[Habit Tracker]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/health-symptom-tracker.md b/wiki/sources/health-symptom-tracker.md new file mode 100644 index 00000000..b2b0984a --- /dev/null +++ b/wiki/sources/health-symptom-tracker.md @@ -0,0 +1,29 @@ +--- +title: "Health & Symptom Tracker" +type: source +tags: [OpenClaw, Health, Symptom, Telegram, Cron] +date: 2026-04-14 +--- + +## Summary +- 核心主题:饮食与症状追踪 +- 问题域:如何通过长期日志发现食物敏感源 +- 方法/机制:Telegram 主题记录食物和症状,定时提醒 + 周分析 +- 结论/价值:模式分析发现潜在诱因 + +## Key Claims +- Telegram 主题消息记录食物和症状,自动附时间戳 +- 每日 3 次定时提醒(早/中/晚)记录饮食 +- 周分析识别食物与症状关联、时间模式、明确诱因 + +## Key Concepts +- [[Health Tracker]]:健康追踪 +- [[Symptom Pattern Analysis]]:症状模式分析 +- [[Food Sensitivity]]:食物敏感识别 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Health Symptom Tracker]] +- [[Telegram]] ← 日志渠道 ← [[Health Symptom Tracker]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/inbox-declutter.md b/wiki/sources/inbox-declutter.md new file mode 100644 index 00000000..c34376fb --- /dev/null +++ b/wiki/sources/inbox-declutter.md @@ -0,0 +1,31 @@ +--- +title: "Inbox De-clutter" +type: source +tags: [OpenClaw, Gmail, Newsletter, Digest, Automation] +sources: ["https://clawhub.ai/kai-jar/gmail-oauth"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:邮件 newsletter 摘要 +- 问题域:如何处理堆积未读的 newsletter +- 方法/机制:Gmail OAuth 读取过去 24 小时 newsletter,生成摘要后询问反馈优化 +- 结论/价值:专用邮箱收集订阅 + AI 筛选精华 + +## Key Claims +- 创建专用 OpenClaw 邮箱接收所有 newsletter +- 每日 8 PM cron 任务读取过去 24 小时 newsletter +- 生成摘要包含精华内容和阅读链接 +- 询问反馈并更新记忆以优化未来筛选 + +## Key Concepts +- [[Inbox De-clutter]]:收件箱清理 +- [[Gmail OAuth]]:Gmail 认证集成 +- [[Newsletter Digest]]:订阅邮件摘要 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Inbox De-clutter]] +- [[Gmail]] ← 数据源 ← [[Inbox De-clutter]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/install-apache-superset-in-docker.md b/wiki/sources/install-apache-superset-in-docker.md new file mode 100644 index 00000000..316e83e0 --- /dev/null +++ b/wiki/sources/install-apache-superset-in-docker.md @@ -0,0 +1,32 @@ +--- +title: "Install Apache Superset in Docker" +type: source +tags: [apache, bi, docker, mysql, superset] +date: 2026-04-13 +source_file: raw/Technical/Home Office/Install Apache Superset in Docker.md +--- + +## Summary +- 核心主题:Apache Superset 在 Docker 环境下的安装与配置 +- 问题域:BI可视化平台自托管部署 +- 方法/机制:Docker镜像部署 + FAB管理员账户创建 + 数据库初始化 +- 结论/价值:通过GHA版本快速部署可用的Superset实例 + +## Key Claims +- Docker pull可直接获取Superset GHA版本镜像 +- 容器端口映射8777:8088实现宿主机访问Superset Web UI +- 必须通过superset fab create-admin创建初始管理员账户 +- load_examples命令加载示例数据集用于功能验证 + +## Key Concepts +- [[Apache Superset]]:开源BI和数据可视化平台,支持SQL查询和图表构建 +- [[Docker容器化部署]]:通过Docker实现应用与环境隔离的部署方式 + +## Key Entities +- [[Docker]]:容器化平台,Superset运行的基础环境 +- [[MySQL]]:Superset默认元数据库后端 + +## Connections +- [[Docker容器化部署]] ← 应用场景 ← [[Apache Superset]] + +## Contradictions diff --git a/wiki/sources/jia-ting-jian-kong-fang-an-prometheus-grafana-node-exporter-cadrvisor-blackbox.md b/wiki/sources/jia-ting-jian-kong-fang-an-prometheus-grafana-node-exporter-cadrvisor-blackbox.md new file mode 100644 index 00000000..56ccdf0c --- /dev/null +++ b/wiki/sources/jia-ting-jian-kong-fang-an-prometheus-grafana-node-exporter-cadrvisor-blackbox.md @@ -0,0 +1,92 @@ +--- +title: "家庭监控方案:Prometheus + Grafana + Node Exporter + cAdvisor + Blackbox" +type: source +tags: [prometheus, grafana, monitoring, node-exporter, cadvisor, blackbox, docker] +date: 2025-11-11 +source_file: raw/Technical/Home Office/🟠家庭监控方案:Prometheus + Grafana + Node Exporter + cAdvisor +Blackbox.md +--- + +## Summary +- 核心主题:家庭环境监控完整解决方案 +- 问题域:主机监控、容器监控、服务可用性监控、日志聚合 +- 方法/机制:Docker Compose一键部署Prometheus+Grafana监控栈 +- 结论/价值:构建工作室级监控体系 + +## Key Claims +- node_exporter采集主机指标(CPU/内存/磁盘/网络) +- cAdvisor采集容器资源使用情况 +- blackbox_exporter探测HTTP/TCP服务可用性 +- Alertmanager支持邮件/Slack/Webhook告警 + +## Key Concepts +- [[监控体系]]:Prometheus + Grafana核心架构 +- [[指标采集]]:Pull模式主动采集exporter数据 +- [[容器监控]]:cAdvisor监控Docker容器 +- [[合成监测]]:blackbox_exporter做主动探测 +- [[告警分发]]:Alertmanager统一告警路由 + +## 架构组件 + +### 数据采集层 +| Exporter | 用途 | 端口 | +|----------|------|------| +| node_exporter | 主机指标 | 9100 | +| cAdvisor | 容器指标 | 8080 | +| blackbox_exporter | HTTP/TCP探测 | 9115 | + +### 存储层 +| 组件 | 用途 | +|------|------| +| Prometheus | 时序数据库/告警规则 | +| Alertmanager | 告警分发 | + +### 展示层 +| 组件 | 用途 | +|------|------| +| Grafana | 可视化仪表盘/告警 | + +## 推荐监控项 +- 主机:CPU使用率>85%告警 +- 磁盘:剩余空间<10%告警 +- 内存:可用内存<15%告警 +- 容器:重启次数>0告警 +- HTTP:探测失败连续2分钟告警 +- TLS:证书剩余<14天告警 + +## Docker Compose部署 +```yaml +services: + prometheus: + image: prom/prometheus:latest + ports: + - "9090:9090" + grafana: + image: grafana/grafana:latest + ports: + - "3000:3000" + node_exporter: + image: prom/node-exporter:latest + network_mode: "host" + cadvisor: + image: gcr.io/cadvisor/cadvisor:latest + ports: + - "8080:8080" + blackbox: + image: prom/blackbox-exporter:latest + ports: + - "9115:9115" +``` + +## Grafana Dashboard ID +| Dashboard | ID | +|-----------|-----| +| Node Exporter Full | 1860 | +| cAdvisor Container Metrics | 14282 | +| Blackbox Exporter Probe | 7587 | + +## 访问地址 +- Prometheus: http://192.168.3.47:9090 +- Grafana: http://192.168.3.47:3000 (admin/admin) +- cAdvisor: http://192.168.3.47:8080 +- node_exporter: http://192.168.3.47:9100/metrics +- blackbox: http://192.168.3.47:9115 \ No newline at end of file diff --git a/wiki/sources/jia-ting-wang-luo-huan-jing-gai-lan-2026-04-03.md b/wiki/sources/jia-ting-wang-luo-huan-jing-gai-lan-2026-04-03.md new file mode 100644 index 00000000..9fd48113 --- /dev/null +++ b/wiki/sources/jia-ting-wang-luo-huan-jing-gai-lan-2026-04-03.md @@ -0,0 +1,87 @@ +--- +title: "家庭网络环境概览 2026-04-03" +type: source +tags: [home-office, nas, synology, ubuntu, vps, network] +date: 2026-04-03 +source_file: raw/Technical/Home Office/🟢家庭网络环境概览_2026-04-03.md +--- + +## Summary +- 核心主题:家庭工作室网络环境完整概览 +- 问题域:多设备管理、服务部署、域名映射 +- 方法/机制:FRP内网穿透 + Caddy反向代理 + Cloudflare DNS +- 结论/价值:构建完整的小型数据中心架构 + +## Key Claims +- Mac Mini M4作为主控节点运行OpenClaw +- Synology NAS DS718运行多个Docker服务 +- 两台Ubuntu Server分别运行不同业务 +- 所有服务通过FRP暴露到公网 + +## Key Entities +- [[Mac Mini]]:主控节点(192.168.3.189) +- [[Synology NAS DS718]]:NAS设备(192.168.3.17) +- [[Ubuntu Server 1]]:应用服务器(192.168.3.47) +- [[Ubuntu Server 2]]:开发服务器(192.168.3.45) +- [[VPS]]:RackNerd公网服务器(192.227.222.142) +- [[OpenClaw]]:AI助手框架 + +## 网络架构 +``` +Internet + ↓ +VPS (RackNerd) - 公网IP + │ + ├── Caddy (HTTPS反向代理) + └── FRPS (端口7000) + │ + ├── Mac Mini (frpc) - 192.168.3.189 + │ └── OpenClaw, vaultwarden, stq + ├── Ubuntu1 (frpc) - 192.168.3.47 + │ └── Grafana, n8n, tiktok_pm + ├── Ubuntu2 (frpc) - 192.168.3.45 + │ └── n8n, gitea, drawio + └── NAS (frpc) - 192.168.3.17 + └── Jellyfin, Navidrome, Calibre +``` + +## 服务列表 + +### Mac Mini M4 +| 服务 | 端口 | 说明 | +|------|------|------| +| OpenClaw | 8080 | AI助手框架 | +| vaultwarden | 5151 | 密码管理 | + +### Synology NAS DS718 +| 服务 | 端口 | 说明 | +|------|------|------| +| DSM | 5000 | 管理系统 | +| Jellyfin | 8096 | 媒体服务器 | +| Navidrome | 4533 | 音乐流媒体 | +| Calibre | 8083 | 电子书库 | +| MinIO | 9001 | 对象存储 | + +### Ubuntu Server 1 +| 服务 | 端口 | 说明 | +|------|------|------| +| Grafana | 3000 | 监控看板 | +| n8n | 62000 | 工作流自动化 | +| Prometheus | 9090 | 监控系统 | +| Superset | 8777 | BI平台 | + +### Ubuntu Server 2 +| 服务 | 端口 | 说明 | +|------|------|------| +| n8n | 5678 | 工作流自动化 | +| Gitea | 3000 | Git服务 | +| drawio | 8085 | 图表编辑 | + +## 域名映射(通过Caddy) +- vaultwarden.ishenwei.online → Mac Mini vaultwarden +- n8n.ishenwei.online → Ubuntu2 n8n +- grafana.ishenwei.online → Ubuntu1 Grafana +- nas.ishenwei.online → NAS DSM + +## 科学上网 +所有服务器配置socks5代理(127.0.0.1:10808) \ No newline at end of file diff --git a/wiki/sources/ke-zi-dong-hua-ke-kuo-zhan-ai-zeng-qiang-de-dian-shang-shu-ju-cai-ji-yu-chu-li-xi-tong.md b/wiki/sources/ke-zi-dong-hua-ke-kuo-zhan-ai-zeng-qiang-de-dian-shang-shu-ju-cai-ji-yu-chu-li-xi-tong.md new file mode 100644 index 00000000..4f0d5ddc --- /dev/null +++ b/wiki/sources/ke-zi-dong-hua-ke-kuo-zhan-ai-zeng-qiang-de-dian-shang-shu-ju-cai-ji-yu-chu-li-xi-tong.md @@ -0,0 +1,54 @@ +--- +title: "可自动化、可扩展、AI增强的电商数据采集与处理系统" +type: source +tags: [scrapy, playwright, n8n, docker, ollama, ecommerce] +date: 2025-11-11 +source_file: raw/Technical/Home Office/可自动化、可扩展、AI增强的电商数据采集与处理系统.md +--- + +## Summary +- 核心主题:基于Docker + Ubuntu + n8n搭建的自动化电商数据采集与AI处理系统 +- 问题域:电商数据采集效率、动态页面处理、AI内容分析 +- 方法/机制:Scrapy + Playwright爬虫组合,n8n工作流自动化,Ollama本地LLM处理 +- 结论/价值:提供完整的电商数据采集、处理、存储、可视化方案 + +## Key Claims +- Scrapy + Playwright组合可解决动态渲染页面抓取问题 +- n8n工作流可实现完整的自动化数据管道 +- 本地Ollama可替代外部API进行AI内容处理 +- Docker容器化部署确保系统可扩展性 + +## Key Concepts +- [[爬虫框架]]:Scrapy高性能爬虫框架 +- [[浏览器渲染]]:Playwright处理JavaScript动态内容 +- [[工作流自动化]]:n8n可视化自动化平台 +- [[本地LLM]]:Ollama本地模型服务 +- [[数据管道]]:ETL数据处理流程 + +## Key Entities +- [[Docker]]:容器化部署基础设施 +- [[n8n]]:工作流自动化平台 +- [[Scrapy]]:Python爬虫框架 +- [[Playwright]]:浏览器自动化工具 +- [[Ollama]]:本地LLM运行平台 + +## Connections +- [[Docker]] ← 承载 ← [[n8n]] +- [[Scrapy]] ← 数据源 ← [[数据管道]] +- [[Ollama]] ← 处理层 ← [[n8n]] + +## Architecture +系统分为三个层次: +1. **数据采集层**:Scrapy/Playwright采集电商数据 +2. **数据处理层**:n8n + LLM API进行清洗、分类、摘要 +3. **存储展示层**:PostgreSQL/MinIO存储,Grafana可视化 + +## 技术栈 +| 组件 | 用途 | +|------|------| +| Scrapy | 结构化页面抓取 | +| scrapy-playwright | 动态页面渲染 | +| n8n | 工作流自动化 | +| Ollama | 本地AI处理 | +| PostgreSQL | 结构化数据存储 | +| MinIO | 对象存储(图片/视频) | \ No newline at end of file diff --git a/wiki/sources/knowledge-base-rag.md b/wiki/sources/knowledge-base-rag.md new file mode 100644 index 00000000..22f339fa --- /dev/null +++ b/wiki/sources/knowledge-base-rag.md @@ -0,0 +1,31 @@ +--- +title: "Personal Knowledge Base (RAG)" +type: source +tags: [OpenClaw, RAG, Knowledge Base, Semantic Search, Telegram] +sources: ["https://clawhub.ai"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:个人知识库(RAG) +- 问题域:如何让收藏的内容可搜索 +- 方法/机制:URL 丢入 Telegram/Slack 自动摄入,语义搜索召回 +- 结论/价值:所有收藏内容语义可查询 + +## Key Claims +- URL 丢入 Telegram/Slack 自动摄入(文章、推文、YouTube 字幕、PDF) +- 语义搜索返回排序结果和来源摘要 +- 可被其他工作流调用(如视频创意工作流查询相关素材) + +## Key Concepts +- [[Personal Knowledge Base]]:个人知识库 +- [[RAG]] ← 底层 ← [[Knowledge Base]] +- [[Semantic Search]]:语义搜索 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Knowledge Base]] +- [[RAG]] ← 技术 ← [[Knowledge Base]] +- [[Telegram]] ← 摄入渠道 ← [[Knowledge Base]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/latex-paper-writing.md b/wiki/sources/latex-paper-writing.md new file mode 100644 index 00000000..5b8f5479 --- /dev/null +++ b/wiki/sources/latex-paper-writing.md @@ -0,0 +1,31 @@ +--- +title: "LaTeX Paper Writing" +type: source +tags: [OpenClaw, LaTeX, Paper Writing, Prismer, Academic] +sources: ["https://github.com/Prismer-AI/Prismer/tree/main/skills/latex-compiler"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:AI 协作 LaTeX 论文写作 +- 问题域:如何无需本地安装 TeX 环境即可编译 LaTeX +- 方法/机制:latex-compiler skill + Prismer Docker 容器提供在线编译 +- 结论/价值:对话式写作 + 即时编译预览 + +## Key Claims +- 对话式协作撰写 LaTeX,描述需求生成源码 +- 无需本地 TeX 安装,在线编译(pdflatex/xelatex/lualatex) +- 内联预览 PDF,支持 IEEE/Article/Beamer/中文模板 +- BibTeX/BibLaTeX 参考文献支持 + +## Key Concepts +- [[LaTeX Writing]]:LaTeX 论文协作写作 +- [[latex-compiler skill]]:LaTeX 编译工具集 +- [[Prismer]] ← 运行环境 ← [[LaTeX Writing]] + +## Connections +- [[OpenClaw]] ← 运行 ← [[LaTeX Writing]] +- [[Prismer]] ← 容器 ← [[LaTeX Writing]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/linux-system-monitor.md b/wiki/sources/linux-system-monitor.md new file mode 100644 index 00000000..7f78fb43 --- /dev/null +++ b/wiki/sources/linux-system-monitor.md @@ -0,0 +1,40 @@ +--- +title: "These 6 Linux apps let you monitor system resources in style" +type: source +tags: [Linux, System-Monitor, TUI] +date: 2025-12-16 +source_file: raw/Technical/These 6 Linux apps let you monitor system resources in style.md +--- + +## Summary +- 核心主题:6款Linux系统资源监控工具推荐 +- 问题域:替代桌面环境自带资源管理器的轻量方案 +- 方法/机制:TUI应用 + GUI应用,按需选择 +- 结论/价值:btop++是TUI首选,功能与美观平衡 + +## Key Claims +- TUI监控应用:响应快、SSH友好 +- btop++:最佳TUI监控,界面美观,功能平衡 +- htop:更聚焦进程的轻量TUI +- Glances:更轻量,完全键盘驱动 +- Bottom:专注实时性能图表 +- Mission Center:全功能GUI,类Task Manager +- Stacer:功能最全的GUI方案,可配置启动项、卸载包、清理缓存 + +## Key Concepts +- [[TUI]]:Text User Interface,终端用户界面 +- [[系统监控]]:CPU/内存/进程/网络监控 + +## Key Entities +- [[btop++]]:最喜欢的TUI监控应用 +- [[htop]]:轻量进程监控 +- [[Glances]]:轻量全览监控 +- [[Bottom]]:实时图表监控 +- [[Mission-Center]]:GUI任务管理器风格 +- [[Stacer]]:功能丰富的GUI系统工具 + +## Connections +- [[Trae远程开发部署指南]] ← 服务器运维 ← [[Linux系统监控工具]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/llms-rag-ai-agent-differences.md b/wiki/sources/llms-rag-ai-agent-differences.md new file mode 100644 index 00000000..56186240 --- /dev/null +++ b/wiki/sources/llms-rag-ai-agent-differences.md @@ -0,0 +1,66 @@ +--- +title: "LLMs、RAG、AI Agent 三个到底什么区别?" +type: source +tags: [llm, rag, ai-agent, ai-basics] +sources: ["https://mp.weixin.qq.com/s/8B_Phrjz_Mlvpe7vJ3maPA"] +last_updated: 2025-11-19 +--- + +## Summary + +- 核心主题:LLMs、RAG、AI Agent三个核心AI概念的定义与区别 +- 问题域:AI应用开发者必须掌握的基础概念澄清 +- 方法/机制:分层解析——LLM是思考、RAG是信息获取、AI Agent是行动执行 +- 结论/价值:三者不是竞争技术,而是在不同层面满足不同场景的能力展示,应结合使用 + +## Key Claims + +- LLM(大语言模型)是AI应用的"天才大脑",但知识有时间节点,无法获取实时信息 +- RAG(检索增强生成)是记忆系统,将LLM链接到外部实时知识库,解决幻觉和知识时效问题 +- AI Agent(智能体)是循环控制系统,感知目标、规划步骤、执行动作、反思结果 +- 真正生产系统需要三者结合:LLM推理+RAG准确性+Agent自主性 + +## Key Quotes + +> "它们并不是竞争技术,而是在三个不同层面,满足不同实际场景的能力展示" — 核心澄清 + +> "LLM在思考方面非常出色,但对当前情况却一无所知" — LLM局限性 + +> "RAG就像是给那个全能天才大脑配备了一位随身图书馆助理" — RAG定义 + +> "用LLM进行推理,用RAG确保准确性,用Agent框架实现自主性" — 三者关系 + +## Key Concepts + +- [[LLM]](Large Language Model,大语言模型):AI应用的天才大脑,擅长思考但知识有时效性 +- [[RAG]](Retrieval-Augmented Generation,检索增强生成):将静态LLM链接到外部实时知识库的记忆系统 +- [[AI Agent]](AI智能体):围绕LLM构建的循环控制系统,感知-规划-执行-反思 +- [[检索增强生成]]:检索(Retrieval)+ 增强生成(Augmented Generation)两步流程 +- [[幻觉问题]]:LLM生成看似合理但实际错误答案的问题 +- [[上下文Context]]:RAG将检索结果作为上下文输入给LLM + +## Key Entities + +- [[ChatGPT]]:底座通用大模型 +- [[DeepSeek]]:底座通用大模型 +- [[Qwen]]:底座通用大模型 +- [[Midjourney]]:专有绘画模型 +- [[Stable Diffusion]]:专有绘画模型 +- [[Claude]]:编程专有模型 + +## Connections + +- [[LLM]] ← 提供者 ← [[思考能力]] +- [[RAG]] ← 提供者 ← [[实时信息]] +- [[AI Agent]] ← 提供者 ← [[行动能力]] +- [[LLM]] ← 扩展 ← [[RAG]] +- [[LLM]] ← 驱动 ← [[AI Agent]] +- [[AI Agent]] ← 依赖 ← [[LLM]] +- [[AI Agent]] ← 依赖 ← [[RAG]] + +## Contradictions + +- 与单一技术万能论冲突: + - 冲突点:是否需要多种技术结合 + - 当前观点:三者结合才能构建完整AI应用 + - 对方观点:单一LLM可以解决所有问题 diff --git a/wiki/sources/local-crm-framework.md b/wiki/sources/local-crm-framework.md new file mode 100644 index 00000000..86245903 --- /dev/null +++ b/wiki/sources/local-crm-framework.md @@ -0,0 +1,32 @@ +--- +title: "Local CRM Framework with DenchClaw" +type: source +tags: [OpenClaw, CRM, DuckDB, DenchClaw, Browser Automation] +sources: ["https://github.com/DenchHQ/DenchClaw"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:本地 CRM 框架 +- 问题域:如何一句话安装完整的本地 CRM +- 方法/机制:DenchClaw 一键安装(DuckDB + Web UI + OpenClaw + 浏览器自动化) +- 结论/价值:一个 npx 命令替代周末配置工作 + +## Key Claims +- 一键安装:npx denchclaw,自动配置完整栈 +- 自然语言 CRM:对象/字段/视图均可用自然语言创建和修改 +- 浏览器自动化:复制 Chrome 配置,Agent 继承相同认证状态 +- 多种视图:Table、Kanban、Calendar、Timeline、Gallery + +## Key Concepts +- [[DenchClaw]]:本地 CRM 框架 +- [[DuckDB]]:嵌入式数据库(无服务器、无凭证) +- [[Browser Automation]]:浏览器自动化(复制 Chrome 配置) +- [[App Builder]]:内置 Web 应用构建器 + +## Connections +- [[OpenClaw]] ← 运行 ← [[DenchClaw CRM]] +- [[DuckDB]] ← 数据库 ← [[DenchClaw CRM]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/mac-mini-an-zhuang-frp-0-65-0-arm64-cao-zuo-bi-ji.md b/wiki/sources/mac-mini-an-zhuang-frp-0-65-0-arm64-cao-zuo-bi-ji.md new file mode 100644 index 00000000..c93370e3 --- /dev/null +++ b/wiki/sources/mac-mini-an-zhuang-frp-0-65-0-arm64-cao-zuo-bi-ji.md @@ -0,0 +1,52 @@ +--- +title: "Mac Mini 安装 FRP 0.65.0(ARM64)操作笔记" +type: source +tags: [frp, mac-mini, arm64, 内网穿透, launchd] +date: 2026-04-13 +source_file: raw/Technical/Home Office/🟣Mac Mini 安装 FRP 0.65.0(ARM64)操作笔记.md +--- + +## Summary +- 核心主题:Mac Mini M4上安装配置FRP内网穿透客户端 +- 问题域:macOS系统软件安装、Gatekeeper权限、内网穿透配置 +- 方法/机制:下载解压 → Gatekeeper解除 → 配置 → launchd自启 +- 结论/价值:实现从公网通过VPS SSH访问内网Mac Mini + +## Key Claims +- FRP 0.65.0支持Apple Silicon ARM64架构 +- macOS需要xattr解除Gatekeeper限制 +- launchd是macOS推荐的开机自启方案 +- tmux/nohup可作为短期后台运行替代方案 + +## Key Concepts +- [[内网穿透]]:通过FRP实现外网访问内网服务 +- [[Gatekeeper]]:macOS应用安全限制机制 +- [[launchd]]:macOS系统服务管理 +- [[FRP]]:高性能内网穿透工具 + +## Key Entities +- [[Mac Mini]]:Apple Silicon主控节点(192.168.3.189) +- [[FRP]]:内网穿透工具 + +## Connections +- [[Mac Mini]] ← 运行 ← [[FRP]] +- [[FRP]] ← 隧道 ← [[VPS]] + +## 配置参数 +| 参数 | 值 | +|------|-----| +| 安装路径 | /opt/frp/frp_0.65.0_darwin_arm64 | +| 服务器地址 | 192.227.222.142 | +| 服务器端口 | 7000 | +| SSH映射 | 22 → 60026 | + +## 后台运行方式 +1. **tmux**(推荐):适合长期运行 +2. **nohup**:简单后台任务 +3. **launchd**(推荐开机自启):系统级服务管理 + +## 端口映射 +| 名称 | 类型 | localPort | remotePort | +|------|------|------------|------------| +| macmini-ssh | tcp | 22 | 60026 | +| vaultwarden | tcp | 5151 | 15151 | \ No newline at end of file diff --git a/wiki/sources/mac-mini-fu-wu-qi-pei-zhi-fang-zhi-zi-dong-suo-ping-yu-shui-mian.md b/wiki/sources/mac-mini-fu-wu-qi-pei-zhi-fang-zhi-zi-dong-suo-ping-yu-shui-mian.md new file mode 100644 index 00000000..2e7542af --- /dev/null +++ b/wiki/sources/mac-mini-fu-wu-qi-pei-zhi-fang-zhi-zi-dong-suo-ping-yu-shui-mian.md @@ -0,0 +1,71 @@ +--- +title: "Mac Mini 服务器配置:防止自动锁屏与睡眠" +type: source +tags: [mac-mini, server, pmset, sleep, caffeinate] +date: 2026-03-15 +source_file: raw/Technical/Home Office/Mac-Mini-服务器配置-防止自动锁屏与睡眠.md +--- + +## Summary +- 核心主题:配置Mac Mini作为无头服务器,防止自动锁屏和睡眠 +- 问题域:服务器长时间运行、远程访问可用性 +- 方法/机制:pmset电源管理 + caffeinate临时保持唤醒 +- 结论/价值:确保远程桌面(如RustDesk)持续可访问 + +## Key Claims +- pmset可禁用所有自动睡眠和待机模式 +- caffeinate适合临时测试而不修改系统设置 +- WOL(网络唤醒)可远程唤醒Mac Mini + +## Key Concepts +- [[电源管理]]:pmset命令控制系统电源行为 +- [[无头服务器]]:无需显示器的服务器配置 +- [[WOL]]:网络唤醒功能 + +## Key Entities +- [[Mac Mini]]:服务器硬件 + +## pmset命令 +```bash +sudo pmset -a sleep 0 # 禁止系统睡眠 +sudo pmset -a displaysleep 0 # 禁止显示器关闭 +sudo pmset -a standby 0 # 禁止待机模式 +sudo pmset -a hibernatemode 0 # 禁止休眠 +sudo pmset -a womp 1 # 启用网络唤醒 +``` + +## 参数说明 +| 参数 | 作用 | +|------|------| +| -a | 应用于所有电源模式 | +| -b | 仅电池模式 | +| -c | 仅电源适配器模式 | + +## caffeinate临时方案 +```bash +# 安装 +brew install caffeinate + +# 保持唤醒(按Ctrl+C停止) +caffeinate -d -i -s +``` + +| 参数 | 作用 | +|------|------| +| -d | 防止显示器睡眠 | +| -i | 防止系统空闲时睡眠 | +| -s | 防止系统睡眠 | +| -u | 模拟用户活动 | + +## 验证当前设置 +```bash +pmset -g # 查看当前电源设置 +pmset -g sleep # 查看睡眠设置 +pmset -g displaysleep # 查看显示器设置 +``` + +## 注意事项 +1. 需要sudo权限 +2. 关闭睡眠会增加功耗(适合接电服务器场景) +3. WOL可远程唤醒设备 +4. 建议设置强密码和防火墙 \ No newline at end of file diff --git a/wiki/sources/macos-chuang-jian-yu-jie-chu-symbolic-link-openclaw-mu-lu-ying-she.md b/wiki/sources/macos-chuang-jian-yu-jie-chu-symbolic-link-openclaw-mu-lu-ying-she.md new file mode 100644 index 00000000..8c39faaf --- /dev/null +++ b/wiki/sources/macos-chuang-jian-yu-jie-chu-symbolic-link-openclaw-mu-lu-ying-she.md @@ -0,0 +1,30 @@ +--- +title: "macOS 创建与解除 Symbolic Link(OpenClaw 目录映射)" +type: source +tags: [macos, symbolic-link, openclaw, obsidian] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟣macOS 创建与解除 Symbolic Link(OpenClaw 目录映射).md +--- + +## Summary +- 核心主题:macOS系统下 Symbolic Link(符号链接)的创建与解除,用于将OpenClaw隐藏目录映射为普通目录 +- 问题域:OpenClaw使用~/.openclaw隐藏目录,不便于在Finder或Obsidian中直接访问 +- 方法/机制:通过ln -s命令创建符号链接,将~/.openclaw映射到~/openclaw +- 结论/价值:实现OpenClaw数据在Obsidian中的直接访问,同时保持OpenClaw兼容性 + +## Key Claims +- Symbolic Link只删除链接文件,不删除真实目录,安全性可控 +- 推荐使用推荐目录结构(agents/skills/memory/prompts/logs/docs)便于Git管理与备份 +- 创建反向链接方案:~/openclaw为实际目录,~/.openclaw指向它 + +## Key Concepts +- [[Symbolic Link]]:符号链接,Unix/Linux系统中指向另一个文件的特殊文件类型 +- [[OpenClaw]]:AI Agent操作系统 +- [[Obsidian]]:本地笔记管理工具 + +## Key Entities +- [[OpenClaw]]:AI Agent操作系统 + +## Connections +- [[OpenClaw]] ← 目录映射 ← [[Symbolic Link]] +- [[Symbolic Link]] ← 兼容 ← [[Obsidian]] diff --git a/wiki/sources/market-research-product-factory.md b/wiki/sources/market-research-product-factory.md new file mode 100644 index 00000000..b6535b38 --- /dev/null +++ b/wiki/sources/market-research-product-factory.md @@ -0,0 +1,29 @@ +--- +title: "Market Research & Product Factory" +type: source +tags: [OpenClaw, Market Research, Reddit, Last 30 Days, Product Building] +sources: ["https://github.com/mvanhorn/last30days-skill/"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:市场调研与产品工厂 +- 问题域:如何从真实用户痛点发现产品机会 +- 方法/机制:Last 30 Days skill 挖掘 Reddit/X 上过去 30 天真实抱怨,AI 直接构建 MVP +- 结论/价值:创业自动化:从问题发现到原型构建 + +## Key Claims +- 研究任意话题 Reddit/X 上过去 30 天的真实挑战和抱怨 +- 从真实用户痛点识别产品机会 +- 让 OpenClaw 直接构建解决痛点的 MVP + +## Key Concepts +- [[Last 30 Days Skill]]:挖掘 Reddit/X 近 30 天内容的工具 +- [[Product Factory]]:从痛点发现到产品构建的完整流水线 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Market Research]] +- [[Last 30 Days Skill]] ← 工具 ← [[Market Research]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/mcp-in-cursor.md b/wiki/sources/mcp-in-cursor.md new file mode 100644 index 00000000..4623469e --- /dev/null +++ b/wiki/sources/mcp-in-cursor.md @@ -0,0 +1,35 @@ +--- +title: "MCP在Cursor中的集成与应用详解" +type: source +tags: [AI, Cursor, MCP, AI-Agent] +date: 2026-04-14 +source_file: raw/Technical/MCP在Cursor中的集成与应用详解.md +--- + +## Summary +- 核心主题:MCP协议在Cursor中的集成和使用方法 +- 问题域:大模型与外围服务的高效集成 +- 方法/机制:MCP Server提供三种接口 → Cursor接入 → Composer Agent模式调用 +- 结论/价值:MCP实现大模型与多样外部工具的无缝链接 + +## Key Claims +- MCP (Modal Context Protocol)是基于Client-Server架构的协议 +- MCP Server提供三种功能接口:资源获取(GET)、工具调用(POST)、Promise提示词 +- Cursor接入MCP方式:SSE服务方式和本地执行命令方式 +- Agent模式自动执行内嵌命令,Normal模式需手动复制执行 + +## Key Concepts +- [[MCP]]:Modal Context Protocol,AI大模型与外围服务集成的协议 +- [[MCP-Server]]:MCP协议体系中的服务提供方 +- [[MCP-Client]]:MCP协议体系中的服务调用方 +- [[SSE]]:Server-Sent Events,一种MCP接入方式 +- [[Sequential-Thinking]]:MCP工具,支持逻辑推理与分步执行任务 + +## Key Entities +- [[Cursor]]:集成了MCP的AI代码编辑器 + +## Connections +- [[Cursor-2-0-指南]] ← MCP支持 ← [[MCP在Cursor中的集成]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/meeting-notes-action-items.md b/wiki/sources/meeting-notes-action-items.md new file mode 100644 index 00000000..6fa35fc3 --- /dev/null +++ b/wiki/sources/meeting-notes-action-items.md @@ -0,0 +1,31 @@ +--- +title: "Automated Meeting Notes & Action Items" +type: source +tags: [OpenClaw, Meeting, Action Items, Jira, Linear, Slack] +date: 2026-04-14 +--- + +## Summary +- 核心主题:会议记录与任务自动提取 +- 问题域:如何让会议结论自动转化为项目任务 +- 方法/机制:转录稿 → 提取决定/行动项 → 自动创建 Jira/Linear/Todoist 任务 +- 结论/价值:会议记录即文档 theater,真正的价值在自动任务创建 + +## Key Claims +- 监听新会议转录稿(Otter.ai/Google Meet/Zoom) +- 提取关键决定、讨论话题、带负责人和截止日期的行动项 +- 自动在 Jira/Linear/Todoist 创建任务并分配 +- 发送 Slack/Discord 摘要,可选截止前提醒 + +## Key Concepts +- [[Meeting Notes]]:会议记录与摘要 +- [[Action Item Extraction]]:行动项自动提取 +- [[Task Creation]]:自动项目任务创建 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Meeting Notes]] +- [[Jira]] ← 任务目标 ← [[Meeting Notes]] +- [[Linear]] ← 任务目标 ← [[Meeting Notes]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/minio-zipline-zi-cheng-guan-tu-chuang-ying-yong-an-zhuang-jiao-cheng.md b/wiki/sources/minio-zipline-zi-cheng-guan-tu-chuang-ying-yong-an-zhuang-jiao-cheng.md new file mode 100644 index 00000000..a5d34208 --- /dev/null +++ b/wiki/sources/minio-zipline-zi-cheng-guan-tu-chuang-ying-yong-an-zhuang-jiao-cheng.md @@ -0,0 +1,40 @@ +--- +title: "MinIO + Zipline 自托管图床应用安装教程" +type: source +tags: [docker, image, minio, n8n, nas, synology, zipline] +date: 2026-04-13 +source_file: raw/Technical/Home Office/🟠MinIO + Zipline 自托管图床应用安装教程.md +--- + +## Summary +- 核心主题:在Synology NAS上通过Docker Compose部署MinIO+S3+Zipline图床 +- 问题域:自托管对象存储与图片托管服务 +- 方法/机制:MinIO提供S3兼容存储 + Zipline提供上传API + PostgreSQL存储元数据 +- 结论/价值:构建可被n8n调用的自托管图床解决方案,支持API Token认证 + +## Key Claims +- Synology NAS上通过Docker Compose部署MinIO、PostgreSQL、Zipline三个容器 +- MinIO bucket必须设置为public-read,否则图片无法直接访问 +- Zipline通过CORE_SECRET环境变量实现安全认证 +- 备份策略采用pg_dump逻辑备份+Hyper Backup增量归档,确保数据库与文件一致性 + +## Key Quotes +> "Zipline → MinIO(S3) → NAS 存储" — 整体架构流向 + +## Key Concepts +- [[S3兼容存储]]:支持S3协议的对象存储系统,MinIO提供兼容实现 +- [[图片托管服务]]:提供图片上传、存储、访问的专用服务 +- [[逻辑备份]]:通过pg_dump导出SQL实现数据库热备份,避免物理备份停机 + +## Key Entities +- [[MinIO]]:S3兼容的对象存储服务器 +- [[Zipline]]:自托管图床应用,支持n8n集成 +- [[Synology-NAS]]:部署的硬件平台 +- [[n8n]]:工作流自动化工具,可调用Zipline API上传图片 + +## Connections +- [[MinIO]] ← 存储层 ← [[Zipline]] +- [[Zipline]] ← 上传接口 ← [[n8n]] +- [[Synology-NAS]] ← 部署平台 ← [[MinIO]] + [[Zipline]] + +## Contradictions diff --git a/wiki/sources/modern-itsm.md b/wiki/sources/modern-itsm.md new file mode 100644 index 00000000..5784ca4a --- /dev/null +++ b/wiki/sources/modern-itsm.md @@ -0,0 +1,38 @@ +--- +title: "Modern ITSM Driving Efficiency, Security & Resilience" +type: source +tags: [ITSM, AIOps, Automation] +sources: [] +last_updated: 2025-03-01 +--- + +## Summary +- 核心主题:现代IT服务管理(ITSM)如何推动效率、安全和韧性 +- 问题域:传统服务管理模型不再可持续,需向敏捷、自动化和韧性转型 +- 方法/机制:AI驱动的异常检测、预测分析、自我修复IT生态系统 +- 结论/价值:ITSM不再是简单的工单系统,而是运营卓越、风险缓解和创新加速的战略推动者 + +## Key Claims +- 问题管理:AI驱动的异常检测和预测分析消除重复故障,ML增强事件关联减少重复工单 +- 事件管理:实时可观测性、自动化修复、AIOps驱动的自我修复IT生态系统将MTTR最小化 +- 变更管理:AI预测失败概率,实现基于风险的变更审批 +- 配置管理:AI驱动的CMDB增强依赖映射、漂移检测、实时影响分析 +- 安全与合规:零信任架构、Policy-as-Code、自动化审计准备 +- 灾难恢复:AI驱动的自动故障转移策略、RTO/RPO优化、DRaaS + +## Key Concepts +- [[ITSM]]:IT Service Management,IT服务管理 +- [[AIOps]]:AI驱动的运维 +- [[CMDB]]:Configuration Management Database,配置管理数据库 +- [[零信任架构]]:Zero Trust Architecture,无隐式信任的安全模型 +- [[Policy-as-Code]]:将安全策略编写为代码实现自动化执行 + +## Key Entities + +## Connections +- [[ITSM]] ← 集成 ← [[AIOps]] +- [[零信任架构]] ← 支撑 ← 安全与合规 +- [[Policy-as-Code]] ← 自动化 ← 合规审计 + +## Contradictions + diff --git a/wiki/sources/multi-account-deployments-cloudformation-stacksets.md b/wiki/sources/multi-account-deployments-cloudformation-stacksets.md new file mode 100644 index 00000000..bcd6c5b5 --- /dev/null +++ b/wiki/sources/multi-account-deployments-cloudformation-stacksets.md @@ -0,0 +1,37 @@ +--- +title: "How to Simplify Multi-Account Deployments Monitoring: Centralized Logs for AWS CloudFormation StackSets" +type: source +tags: [AWS, CloudFormation, Multi-Account, Monitoring] +sources: [] +last_updated: 2025-10-25 +--- + +## Summary +- 核心主题:通过集中式日志监控简化AWS多账户部署 +- 问题域:跨多账户部署CloudFormation StackSets时的运营可见性挑战 +- 方法/机制:利用EventBridge和CloudWatch Logs实现跨账户日志集中收集 +- 结论/价值:集中式日志系统提供跨所有账户的统一监控视图,简化故障排除 + +## Key Claims +- AWS CloudFormation StackSets支持跨多个账户和区域部署基础设施 +- 解决方案架构包含四个组件:管理账户设置、目标账户配置、资源部署、监控可视化 +- 事件流:CloudFormation生成事件→EventBridge规则捕获→跨账户转发→CloudWatch Log Group集中存储 +- CloudWatch Logs Insights可自定义查询实现跨组织CloudFormation事件分析 + +## Key Concepts +- [[AWS CloudFormation StackSets]]:跨多个账户和区域部署基础设施的服务 +- [[Amazon EventBridge]]:无服务器事件总线服务 +- [[Amazon CloudWatch Logs]]:云监控日志服务 +- [[IaC]]:Infrastructure as Code,基础设施即代码 + +## Key Entities +- [[AWS]]:亚马逊云服务提供商 +- [[AWS Organizations]]:AWS组织管理服务 + +## Connections +- [[AWS CloudFormation StackSets]] ← 依赖 ← [[Amazon EventBridge]] +- [[Amazon EventBridge]] ← 转发 ← [[Amazon CloudWatch Logs]] +- [[IaC]] ← 实现 ← CloudFormation + +## Contradictions + diff --git a/wiki/sources/multi-agent-system-reliability.md b/wiki/sources/multi-agent-system-reliability.md new file mode 100644 index 00000000..33f07a77 --- /dev/null +++ b/wiki/sources/multi-agent-system-reliability.md @@ -0,0 +1,62 @@ +--- +title: "Multi-Agent System Reliability" +type: source +tags: [multi-agent, system-reliability, llm, architecture] +sources: ["https://blog.alexewerlof.com/p/multi-agent-system-reliability"] +last_updated: 2026-04-13 +--- + +## Summary + +- 核心主题:四种架构模式提升多智能体系统可靠性 +- 问题域:LLM作为不可靠组件,如何构建企业级可靠多智能体系统 +- 方法/机制:Hierarchy(层级)、Consensus(共识)、Adversarial Debate(对抗式辩论)、Knock-out(淘汰制) +- 结论/价值:将LLM视为分布式系统中不可靠组件,而非魔法聊天机器人 + +## Key Claims + +- 人类系统协作的四种模式可应用于多智能体架构:Hierarchy、Consensus、Adversarial debate、Knock-out +- 不要将LLM拟人化——它不会死亡或挨饿,无法真正感受恐惧 +- 对LLM的"威胁"之所以有效,是因为训练数据中高风险情境产生高质量文本 +- 构建稳健系统需要停止要求模型"小心",而要强制它正确 +- 共识模式:3个模型同时幻觉相同谎言的概率仅为0.8%(0.2^3) + +## Key Quotes + +> "LLMs are slow and error prone. So are human beings. Somehow we manage to build more reliable systems like an army, a company, or a state nation." — 人类与LLM类比 + +> "Don't anthropomorphize LLMs! Find a way to piggy back on their human-corpus training while being aware of their non-biological differences." — 核心忠告 + +> "We don't need AI that 'cares.' We need AI that is constrained, verified, pruned, and challenged." — 可靠性工程思维 + +## Key Concepts + +- [[Multi-Agent System]](多智能体系统):通过并行处理和专业智能体克服LLM速度和通用性局限 +- [[Hierarchy Pattern]](层级模式):Supervisor规划器分配任务给Worker执行,Validator验证结果 +- [[Consensus Pattern]](共识模式):多模型投票,真相从多数意见中浮现 +- [[Adversarial Debate Pattern]](对抗式辩论模式):Generator提出、Critic攻击、Judge裁决 +- [[Knock-out Pattern]](淘汰制模式):适者生存,删除表现最差的代理 +- [[LLM Hallucination]](LLM幻觉):模型生成看似合理但错误的输出 +- [[Context Drift]](上下文漂移):模型在长对话中失去焦点的现象 +- [[Sycophancy]](谄媚):模型过度迎合用户而撒谎的问题 + +## Key Entities + +- [[Alex Ewerlöf]]:作者,资深工程师,可靠性工程专家 +- [[KTH]]:瑞典皇家理工学院,作者学位来源 + +## Connections + +- [[Multi-Agent System]] ← 模式1 ← [[Hierarchy Pattern]] +- [[Multi-Agent System]] ← 模式2 ← [[Consensus Pattern]] +- [[Multi-Agent System]] ← 模式3 ← [[Adversarial Debate Pattern]] +- [[Multi-Agent System]] ← 模式4 ← [[Knock-out Pattern]] +- [[Consensus Pattern]] ← 公式 ← [[LLM Hallucination概率计算]] +- [[Hierarchy Pattern]] ← 包含 ← [[Planner Worker Validator]] + +## Contradictions + +- 与单一LLM万能论冲突: + - 冲突点:LLM可靠性和企业级应用 + - 当前观点:LLM不可靠,需要多层验证机制 + - 对方观点:可通过提示工程解决可靠性问题 diff --git a/wiki/sources/multi-agent-team.md b/wiki/sources/multi-agent-team.md new file mode 100644 index 00000000..22865f6d --- /dev/null +++ b/wiki/sources/multi-agent-team.md @@ -0,0 +1,34 @@ +--- +title: "Multi-Agent Specialized Team (Solo Founder Setup)" +type: source +tags: [OpenClaw, Multi-Agent, Solo Founder, Telegram, Specialized Agents] +sources: ["https://x.com/iamtrebuh/status/2011260468975771862"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:单人创业者多 Agent 专业团队 +- 问题域:如何用多个专业 Agent 替代团队 +- 方法/机制:每个 Agent 有独立角色/人格/模型,通过 Telegram 标签路由,共享记忆 +- 结论/价值:24/7 专业团队,成本接近为零 + +## Key Claims +- 专业 Agent(Milo 策略、Josh 商业、营销、开发) +- 共享记忆 + 私有上下文 +- 单一控制平面:Telegram 群组 tag 路由 +- 定时主动任务:内容提示、竞品监控、指标追踪 +- 并行执行:多个 Agent 同时处理独立任务 + +## Key Concepts +- [[Multi-Agent Team]]:多 Agent 协作团队 +- [[Solo Founder Setup]]:单人创业者配置 +- [[Shared Memory]]:跨 Agent 共享记忆 +- [[Telegram Routing]]:Telegram 标签路由 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Multi-Agent Team]] +- [[Telegram]] ← 控制平面 ← [[Multi-Agent Team]] +- [[Subagent]] ← 实现 ← [[Multi-Agent Team]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/multi-channel-assistant.md b/wiki/sources/multi-channel-assistant.md new file mode 100644 index 00000000..345cbaa9 --- /dev/null +++ b/wiki/sources/multi-channel-assistant.md @@ -0,0 +1,32 @@ +--- +title: "Multi-Channel Personal Assistant" +type: source +tags: [OpenClaw, Multi-Channel, Telegram, Slack, Google Workspace, Todoist] +date: 2026-04-14 +--- + +## Summary +- 核心主题:多渠道统一助手 +- 问题域:如何在一个界面路由到所有工具 +- 方法/机制:Telegram 主题路由 + Slack/Google Workspace/Todoist/Asana 集成 +- 结论/价值:单一对话界面管理所有工作流 + +## Key Claims +- Telegram 主题路由(视频创意/CRM/财报/配置等) +- Slack 团队协作(任务分配、知识库、视频创意触发) +- Google Workspace:日历/邮件/Drive +- Todoist 快速任务捕获、Asana 项目管理 +- 自动化提醒:倒垃圾、周报等 + +## Key Concepts +- [[Multi-Channel Assistant]]:多渠道统一助手 +- [[Telegram Topic Routing]]:Telegram 主题路由 +- [[Google Workspace]]:Google 工作区集成 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Multi-Channel Assistant]] +- [[Telegram]] ← 主界面 ← [[Multi-Channel Assistant]] +- [[Slack]] ← 协作 ← [[Multi-Channel Assistant]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/multi-channel-customer-service.md b/wiki/sources/multi-channel-customer-service.md new file mode 100644 index 00000000..897f609d --- /dev/null +++ b/wiki/sources/multi-channel-customer-service.md @@ -0,0 +1,31 @@ +--- +title: "Multi-Channel AI Customer Service Platform" +type: source +tags: [OpenClaw, Customer Service, WhatsApp, Instagram, Gmail, Google Reviews] +date: 2026-04-14 +--- + +## Summary +- 核心主题:多渠道 AI 客服平台 +- 问题域:小型企业如何 24/7 处理多平台客户咨询 +- 方法/机制:统一收件箱 + AI 自动回复 + 人工升级 +- 结论/价值:80% 咨询自动处理,响应时间从 4 小时降到 2 分钟 + +## Key Claims +- 统一收件箱:WhatsApp Business/Instagram DMs/Gmail/Google Reviews +- AI 自动处理 FAQ、预约请求 +- 人工升级:复杂问题标记审核 +- 测试模式:演示系统不影响真实客户 + +## Key Concepts +- [[Customer Service Platform]]:客服平台 +- [[Multi-Channel Inbox]]:多渠道统一收件箱 +- [[Human Handoff]]:人工升级机制 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Customer Service]] +- [[WhatsApp]] ← 渠道 ← [[Customer Service]] +- [[Instagram]] ← 渠道 ← [[Customer Service]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/multi-cloud-strategy-roi.md b/wiki/sources/multi-cloud-strategy-roi.md new file mode 100644 index 00000000..d671cd92 --- /dev/null +++ b/wiki/sources/multi-cloud-strategy-roi.md @@ -0,0 +1,40 @@ +--- +title: "How Can a Multi Cloud Strategy Transform Your Business ROI" +type: source +tags: [Multi-Cloud, ROI, Cloud Strategy] +sources: [] +last_updated: 2025-03-01 +--- + +## Summary +- 核心主题:多云策略如何提升业务ROI +- 问题域:避免供应商锁定、增加弹性、优化成本、满足合规要求 +- 方法/机制:利用不同云提供商的差异化优势(AWS/Azure/GCP) +- 结论/价值:多云策略通过灵活性、成本优化和风险分散最大化云投资回报 + +## Key Claims +- 78%采用多云策略的企业使用超过3个公有云以提升敏捷性和成本节约 +- 86%公司到2024年底计划采用多云方法 +- 多云策略优化后可实现30%运营成本降低 +- 多云核心价值:避免供应商锁定、增加弹性和可靠性、改善安全态势、可扩展性、成本优化、访问创新、合规灵活性、性能优化 + +## Key Concepts +- [[多云策略]]:使用多个云服务提供商(而非单一供应商)的战略方法 +- [[供应商锁定]]:依赖单一云提供商导致的迁移困难和议价能力丧失 +- [[云成本优化]]:通过资源优化和定价模型优化降低云支出 +- [[数据主权]]:数据存储和处理受地域法规约束 + +## Key Entities +- [[AWS]]:亚马逊云服务 +- [[Azure]]:微软云服务 +- [[Google Cloud]]:谷歌云服务 +- [[Kubernetes]]:容器编排工具 +- [[Terraform]]:IaC工具 + +## Connections +- [[多云策略]] ← 依赖 ← [[Kubernetes]] +- [[多云策略]] ← 依赖 ← [[Terraform]] +- [[云成本优化]] ← 实现 ← 资源优化 + +## Contradictions + diff --git a/wiki/sources/multi-source-tech-news-digest.md b/wiki/sources/multi-source-tech-news-digest.md new file mode 100644 index 00000000..62e1dd07 --- /dev/null +++ b/wiki/sources/multi-source-tech-news-digest.md @@ -0,0 +1,30 @@ +--- +title: "Multi-Source Tech News Digest" +type: source +tags: [OpenClaw, Tech News, RSS, Twitter, GitHub, Discord] +sources: ["https://github.com/draco-agent/tech-news-digest"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:多源科技新闻聚合与评分 +- 问题域:如何从 109+ 来源获取高质量科技新闻 +- 方法/机制:RSS(46) + Twitter KOL(44) + GitHub(19) + 网页搜索(4) 四层管道 +- 结论/价值:自动去重评分,个性化投递 Discord/邮件/Telegram + +## Key Claims +- RSS Feeds(46源)、Twitter KOL(44账号)、GitHub Releases(19仓库)、网页搜索(4主题) +- 去重 + 质量评分(优先级来源+3,多源+5,时效性+2,参与度+1) +- 完全可定制:30 秒添加新来源 + +## Key Concepts +- [[Tech News Digest]]:科技新闻摘要 +- [[Multi-Source Aggregation]]:多源聚合 +- [[RSS Feed]]:RSS 订阅源 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Tech News Digest]] +- [[Discord]] ← 投递 ← [[Tech News Digest]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/multiple-interests-guide.md b/wiki/sources/multiple-interests-guide.md new file mode 100644 index 00000000..b7f59d5a --- /dev/null +++ b/wiki/sources/multiple-interests-guide.md @@ -0,0 +1,68 @@ +--- +title: "If you have multiple interests, do not waste the next 2-3 years" +type: source +tags: [personal-development, creativity, entrepreneurship] +sources: ["https://letters.thedankoe.com/"] +last_updated: 2026-04-14 +--- + +## Summary + +- 核心主题:拥有多重兴趣是超能力,如何将其转化为有利可图的事业 +- 问题域:拥有多种兴趣的人群如何在现代社会中建立独特竞争优势 +- 方法/机制:通过自我教育、自利、自给自足三要素,将兴趣转化为产品、内容和品牌 +- 结论/价值:第二次文艺复兴时期,多重兴趣者应成为创造者而非单一专才 + +## Key Claims + +- 工业时代专业化使人愚蠢依赖,通才型人在当下反而拥有独特优势 +- 个人成功三要素:Self-education(自学)、Self-interest(自利)、Self-sufficiency(自给自足) +- 关注(Attention)是最后的护城河,每个企业都是媒体企业 +- 品牌是人们寻求转变的环境,品牌即内容 +- 系统经济时代,产品即系统 + +## Key Quotes + +> "We don't live in the Industrial Age anymore. Specializing in one skill is almost certain death." — 核心观点 + +> "Your curiosity and love for learning are your advantages in today's world." — 多重兴趣价值 + +> "The truly selfish person...is a self-respecting, self-supporting human being." — 安·兰德观点引用 + +> "Study the science of art. Study the art of science. Develop your senses—especially learn how to see. Realize that everything connects to everything else." — 达·芬奇 + +## Key Concepts + +- [[通才型思维]]:跨领域知识组合创造独特世界观 +- [[自我教育]]:自主驱动学习,而非依赖传统教育体系 +- [[自利]]:关注自身利益,而非服务于组织 +- [[自给自足]]:拒绝外包判断、学习和自主性 +- [[第二次文艺复兴]]:AI时代知识获取成本趋零,个人可追求多领域精通 +- [[注意力护城河]]:在AI可生成任何内容的时代,关注度是最后竞争优势 +- [[品牌环境]]:品牌是用户寻求转变的小世界 +- [[创意密度]]:高质量想法的密集程度决定内容价值 + +## Key Entities + +- [[Adam Smith]]:引用其关于专业化局限性的观点 +- [[达·芬奇]]:文艺复兴通才代表 +- [[米开朗基罗]]:文艺复兴通才代表 +- [[安·兰德]]:引用其关于自私的观点 +- [[Dan Koe]]:文章作者,内容创业者 +- [[Jordan Peterson]]:内容创作者典范 + +## Connections + +- [[通才型思维]] ← 来源 ← [[第二次文艺复兴]] +- [[自我教育]] ← 驱动 ← [[自利]] +- [[自利]] ← 澄清 ← [[自给自足]] +- [[自给自足]] ← 支撑 ← [[通才型思维]] +- [[注意力护城河]] ← 重要性 ← [[品牌环境]] +- [[品牌环境]] ← 构成 ← [[创意密度]] + +## Contradictions + +- 与传统职业发展观冲突: + - 冲突点:专业化vs多元化 + - 当前观点:多兴趣是超专业化的竞争优势 + - 对方观点:专业化是工业时代的成功路径 diff --git a/wiki/sources/mysql-mariadb-shu-ju-ku-xiang-xi-xin-xi.md b/wiki/sources/mysql-mariadb-shu-ju-ku-xiang-xi-xin-xi.md new file mode 100644 index 00000000..bccae72c --- /dev/null +++ b/wiki/sources/mysql-mariadb-shu-ju-ku-xiang-xi-xin-xi.md @@ -0,0 +1,32 @@ +--- +title: "MySQL MariaDB 数据库详细信息" +type: source +tags: [database, mariadb, mysql, nas] +date: 2026-04-13 +source_file: raw/Technical/Home Office/🟠MySQL MariaDB 数据库详细信息.md +--- + +## Summary +- 核心主题:MariaDB数据库的远程访问配置与用户创建 +- 问题域:数据库网络访问安全与权限管理 +- 方法/机制:socket本地登录 + CREATE USER远程授权 + GRANT ALL PRIVILEGES +- 结论/价值:解决新安装MariaDB默认只允许localhost访问的问题 + +## Key Claims +- MariaDB新安装后只有root@localhost,没有root@%或其他远程用户 +- 远程访问失败的根本原因是缺少Host/User组合和对应权限 +- 必须通过CREATE USER 'shenwei'@'%'创建允许任意主机连接的用户 +- FLUSH PRIVILEGES使权限更改立即生效 + +## Key Concepts +- [[数据库远程访问]]:通过TCP/IP网络连接数据库,而非socket本地连接 +- [[MariaDB权限管理]]:基于Host+User组合的访问控制机制 + +## Key Entities +- [[MySQL]]:此处指MariaDB分支,兼容MySQL协议 +- [[Synology-NAS]]:数据库运行的宿主环境 + +## Connections +- [[MySQL]] ← 运行于 ← [[Synology-NAS]] + +## Contradictions diff --git a/wiki/sources/n8n-Claude-通过自然语言自动化工作流.md b/wiki/sources/n8n-Claude-通过自然语言自动化工作流.md new file mode 100644 index 00000000..f4c983b2 --- /dev/null +++ b/wiki/sources/n8n-Claude-通过自然语言自动化工作流.md @@ -0,0 +1,40 @@ +--- +title: "n8n+Claude 通过自然语言自动化工作流" +type: source +tags: [n8n, claude, workflow, MCP] +sources: [] +last_updated: 2026-04-14 +source_file: raw/Technical/Workflow/n8n+Claude 通过自然语言自动化工作流.md +--- + +## Summary +- 核心主题:n8n MCP 项目使 Claude 能通过自然语言自动生成 n8n 工作流 +- 问题域:新手在架构设计和节点选择中的困惑 +- 方法/机制:n8n-mcp 作为桥梁,连接 n8n 工作流平台与 AI 模型,提供 543 个节点的结构化访问 +- 结论/价值:Claude 可自动生成约 80%-90% 正确的工作流布局和逻辑,显著降低学习门槛 + +## Key Claims +- n8n-mcp 提供 543 个 n8n 节点的结构化访问,99% 属性覆盖 +- Claude 通过 n8n-mcp 可理解并使用 n8n 节点 +- 环境搭建需要 Node.js 和 Claude 桌面客户端 +- Claude 生成的工作流约有 10%-20% 错误率,需人工修正 +- 推荐使用 Opensea 模型和开启 extended thinking 模式 + +## Key Quotes +> "n8n-MCP serves as a bridge between n8n's workflow automation platform and AI models, enabling them to understand and work with n8n nodes effectively." + +## Key Concepts +- [[n8n-mcp]]:n8n MCP 服务器,连接 n8n 与 AI 模型 +- [[自然语言工作流生成]]:通过自然语言提示直接生成工作流 +- [[Extended Thinking]]:Claude 深度推理模式 + +## Key Entities +- [[n8n]]:工作流自动化工具 +- [[Claude]]:AI 助手工具 +- [[Node.js]]:JavaScript 运行时环境 +- [[czlonkowski]]:n8n-mcp 项目作者 + +## Connections +- [[Claude]] ← 使用 ← [[n8n-mcp]] ← 连接 ← [[n8n]] +- [[n8n-mcp]] ← 基于 ← [[Node.js]] +- [[自然语言工作流生成]] ← 使用 ← [[Extended Thinking]] \ No newline at end of file diff --git a/wiki/sources/n8n-configure-telegram-trigger.md b/wiki/sources/n8n-configure-telegram-trigger.md new file mode 100644 index 00000000..20af6974 --- /dev/null +++ b/wiki/sources/n8n-configure-telegram-trigger.md @@ -0,0 +1,30 @@ +--- +title: "n8n configure telegram trigger" +type: source +tags: [n8n, telegram] +sources: [] +last_updated: 2026-04-14 +source_file: raw/Technical/Workflow/n8n configure telegram trigger.md +--- + +## Summary +- 核心主题:n8n Telegram Trigger 配置问题解决 +- 问题域:Telegram Webhook 配置需要 HTTPS URL +- 方法/机制:设置 WEBHOOK_URL 环境变量为 HTTPS 地址 +- 结论/价值:通过配置环境变量解决 Telegram Trigger 的 Bad Request 错误 + +## Key Claims +- n8n 实例必须通过 HTTPS URL 公开访问,否则 Telegram Webhook 设置会失败 +- 设置 WEBHOOK_URL 环境变量可让 n8n 生成 HTTPS 的 Webhook URL + +## Key Concepts +- [[Telegram Trigger]]:n8n 中用于接收 Telegram 消息的触发节点 +- [[Webhook HTTPS]]:Telegram 要求 Webhook URL 必须为 HTTPS + +## Key Entities +- [[n8n]]:工作流自动化平台 +- [[Cpolar]]:内网穿透工具,用于暴露本地服务 + +## Connections +- [[n8n]] ← 配置 ← [[Webhook HTTPS]] +- [[n8n]] ← 使用 ← [[Cpolar]] \ No newline at end of file diff --git a/wiki/sources/n8n-docker-install-update.md b/wiki/sources/n8n-docker-install-update.md new file mode 100644 index 00000000..ac93cc6f --- /dev/null +++ b/wiki/sources/n8n-docker-install-update.md @@ -0,0 +1,37 @@ +--- +title: "n8n docker install & update" +type: source +tags: [n8n, docker, workflow] +sources: [] +last_updated: 2026-04-14 +source_file: raw/Technical/Workflow/n8n docker install & update.md +--- + +## Summary +- 核心主题:n8n Docker 安装、网络代理配置与更新流程 +- 问题域:本地部署 n8n 需解决 HTTPS 配置和网络代理问题 +- 方法/机制:Docker Compose + Dockerfile 环境变量配置,通过 ALL_PROXY 设置 SOCKS5 代理 +- 结论/价值:完成 n8n 容器化部署,实现网络代理穿透和自动更新 + +## Key Claims +- n8n 需要配置 N8N_PROTOCOL=https、N8N_HOST 和 WEBHOOK_URL 环境变量 +- 容器内网络代理通过 ALL_PROXY=socks5://172.21.0.1:10808 配置 +- Docker 网桥 IP 需要通过 docker network inspect 确认 +- 防火墙需允许 Docker 网桥访问代理端口 + +## Key Concepts +- [[Docker Compose]]:多容器 Docker 应用定义工具 +- [[环境变量配置]]:N8N_PROTOCOL、WEBHOOK_URL 等 n8n 必要配置 +- [[SOCKS5代理]]:容器内网络代理配置方式 +- [[Caddy]]:反向代理服务器,用于 HTTPS 访问 + +## Key Entities +- [[n8n]]:工作流自动化平台 +- [[Docker]]:容器化平台 +- [[Caddy]]:反向代理/HTTPS 服务器 +- [[V2Ray/Tuic]]:代理工具 + +## Connections +- [[n8n]] ← 运行在 ← [[Docker]] +- [[Docker]] ← 网络通过 ← [[SOCKS5代理]] +- [[n8n]] ← 访问 ← [[Caddy]] ← 提供 ← [[HTTPS]] \ No newline at end of file diff --git a/wiki/sources/n8n-full-tutorial-building-ai-agents-in-2025-for-beginners.md b/wiki/sources/n8n-full-tutorial-building-ai-agents-in-2025-for-beginners.md new file mode 100644 index 00000000..cdb5b6ae --- /dev/null +++ b/wiki/sources/n8n-full-tutorial-building-ai-agents-in-2025-for-beginners.md @@ -0,0 +1,39 @@ +--- +title: "n8n full tutorial building AI agents in 2025 for Beginners!" +type: source +tags: [n8n, ai-agent, tutorial] +sources: [] +last_updated: 2026-04-14 +source_file: raw/Technical/Workflow/n8n full tutorial building AI agents in 2025 for Beginners!.md +--- + +## Summary +- 核心主题:n8n 平台构建 AI Agent 完整教程 +- 问题域:新手学习 AI Agent 构建的入门路径 +- 方法/机制:使用 n8n 的 5 种节点类型(触发器、动作、工具、代码、AI Agent)构建工作流 +- 结论/价值:Agentic System 结合工作流的确定性与 Agent 的灵活性,实现动态任务执行 + +## Key Claims +- Agentic System 由 Agent 和 Workflow 组成,Agent 基于 LLM 动态选择工具 +- n8n 界面直观,通过分类节点降低学习曲线 +- 5 种节点类型(触发器、动作、工具、代码、AI Agent)是构建健壮自动化的基础 +- 内存模块使 Agent 能保留上下文,提升对话连贯性 +- Airtable 等外部工具集成扩展 Agent 能力 + +## Key Quotes +> "Agentic systems combine the predictability of workflows with the flexibility of agents, enabling systems that can adapt to user needs dynamically." + +## Key Concepts +- [[Agentic System]]:代理系统,结合 Agent 和 Workflow 的混合架构 +- [[n8n节点类型]]:触发器、动作、工具、代码、AI Agent +- [[Memory Module]]:内存模块,用于保留对话上下文 +- [[Airtable集成]]:外部数据库集成,用于库存管理 + +## Key Entities +- [[n8n]]:工作流自动化平台 +- [[Airtable]]:在线数据库服务 + +## Connections +- [[n8n]] ← 构建 ← [[Agentic System]] +- [[Agentic System]] ← 使用 ← [[Memory Module]] +- [[Agentic System]] ← 集成 ← [[Airtable]] \ No newline at end of file diff --git a/wiki/sources/n8n-workflow-orchestration.md b/wiki/sources/n8n-workflow-orchestration.md new file mode 100644 index 00000000..27639655 --- /dev/null +++ b/wiki/sources/n8n-workflow-orchestration.md @@ -0,0 +1,31 @@ +--- +title: "OpenClaw + n8n Workflow Orchestration" +type: source +tags: [OpenClaw, n8n, Workflow, Webhook, Security] +sources: ["https://github.com/caprihan/openclaw-n8n-stack"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:OpenClaw + n8n 工作流编排 +- 问题域:如何避免 AI Agent 直接处理 API 凭证 +- 方法/机制:Agent 设计 n8n 工作流,凭证隔离在 n8n 中,Agent 只调用 webhook +- 结论/价值:可观测性(视觉调试)+ 安全性(凭证隔离)+ 性能(确定性任务不走 LLM) + +## Key Claims +- 代理模式:OpenClaw 写 n8n 工作流,凭证留在 n8n +- 凭证隔离:API keys 在 n8n,Agent 只有 webhook URL +- 视觉调试:n8n UI 拖拽查看 +- 锁定工作流:构建测试后锁定,防止 Agent 修改 + +## Key Concepts +- [[n8n Workflow Orchestration]]:n8n 工作流编排 +- [[Webhook Proxy]]:webhook 代理模式 +- [[Credential Isolation]]:凭证隔离 + +## Connections +- [[OpenClaw]] ← 设计者 ← [[n8n Orchestration]] +- [[n8n]] ← 执行者 ← [[n8n Orchestration]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/nano-banana-pro-ti-shi-ci-zhi-nan.md b/wiki/sources/nano-banana-pro-ti-shi-ci-zhi-nan.md new file mode 100644 index 00000000..1ace41d3 --- /dev/null +++ b/wiki/sources/nano-banana-pro-ti-shi-ci-zhi-nan.md @@ -0,0 +1,35 @@ +--- +title: "Nano Banana Pro提示词指南" +type: source +tags: [nano-banana-pro, prompt-engineering, google, image-generation] +date: 2025-12-18 +source_file: raw/AI/谷歌深夜甩出一份【Nano Banana Pro提示词指南】,手把手教你生产专业级内容,实战案例+提示词模版.md +--- + +## Summary +- 核心主题:Nano Banana Pro专业内容生成提示词指南 +- 问题域:AI图像生成、提示词工程 +- 方法/机制:意图理解、物理规则推演、构图美学 +- 结论/价值:4K级专业产线级别的内容生产方案 + +## Key Claims +- Nano Banana Pro具备意图理解引擎突破,不再是简单关键词匹配 +- 支持物理规则推演(光影反射)、构图美学(黄金分割)、语义上下文推理 +- 最多支持14张参考图像,实现身份锁定 +- 利用Google搜索实现实时信息锚定,减少幻觉 + +## Key Quotes +> "停止使用标签堆砌,开始像创意总监一样思考" — 核心原则 +> "编辑,而非重新生成" — 最佳实践 + +## Key Concepts +- [[身份锁定]]:保持角色面部特征一致性的技术 +- [[意图理解引擎]]:理解创作意图而非匹配关键词 +- [[信息锚定]]:基于实时搜索数据减少AI幻觉 + +## Key Entities +- [[Nano Banana Pro]]:Google的图像生成模型 + +## Connections +- [[Nano Banana Pro]] ← 专业内容生成 ← [[意图理解引擎]] +- [[Nano Banana Pro]] ← 信息锚定 ← [[Google搜索]] \ No newline at end of file diff --git a/wiki/sources/nano-banana-ti-shi-ci-kuang-jia.md b/wiki/sources/nano-banana-ti-shi-ci-kuang-jia.md new file mode 100644 index 00000000..70cecd2e --- /dev/null +++ b/wiki/sources/nano-banana-ti-shi-ci-kuang-jia.md @@ -0,0 +1,29 @@ +--- +title: "Nano Banana 提示词框架" +type: source +tags: [nano-banana, 提示词, prompt, AI绘图] +date: 2026-04-14 +--- + +## Summary +- 核心主题:Nano Banana结构化提示词框架模板 +- 问题域:如何标准化AI绘图提示词结构 +- 方法/机制:提供物件描述和人物描述两套JSON框架 +- 结论/价值:为AI绘图提示词提供可复用的结构化模板 + +## Key Claims +- 物件描述框架:shot、subject(item/materials/details/condition)、environment、lighting、camera、color_grade、style、quality、negatives +- 人物描述框架:shot、subject(age/appearance/pose)、environment、lighting、camera、color_grade、style、quality、negatives +- camera参数包含:focal_length、aperture、angle +- negatives字段用于排除不需要的元素 + +## Key Concepts +- [[Nano Banana]] ← 框架 ← [[提示词框架]] +- [[提示词框架]] ← 类型 ← [[物件描述框架]] +- [[提示词框架]] ← 类型 ← [[人物描述框架]] + +## Connections +- [[Nano Banana]] ← 来源 ← [[Nano Banana提示词框架]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/never-write-another-prompt.md b/wiki/sources/never-write-another-prompt.md new file mode 100644 index 00000000..d4e0fd09 --- /dev/null +++ b/wiki/sources/never-write-another-prompt.md @@ -0,0 +1,29 @@ +--- +title: "Never write another prompt" +type: source +tags: [提示词, prompt工程, AI工具] +sources: ["https://youtu.be/OkaplCDf7Ac"] +date: 2025-03-06 +--- + +## Summary +- 核心主题:提示词生成工具简化Prompt Engineering流程 +- 问题域:如何低成本生成高质量结构化提示词 +- 方法/机制:通过描述自动生成详细提示词,支持变量自定义 +- 结论/价值:提示词民主化,让非技术用户也能高效利用AI + +## Key Claims +- 提示词工程民主化:工具自动生成详细提示词,无需专业背景 +- 成本效益:专业提示词服务100-500美元,工具可无限生成 +- 变量支持:通过变量实现提示词复用和定制 +- 提示词库:提供现成提示词参考和灵感 + +## Key Concepts +- [[提示词工程]] ← 简化 ← [[提示词生成工具]] +- [[变量]] ← 用于 ← [[提示词复用]] + +## Connections +- [[提示词优化]] ← 相关 ← [[提示词工程]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/nodewarden-ba-bitwarden-ban-dao-cloudflare-workers-shang-che-di-gao-bie-fu-wu-qi.md b/wiki/sources/nodewarden-ba-bitwarden-ban-dao-cloudflare-workers-shang-che-di-gao-bie-fu-wu-qi.md new file mode 100644 index 00000000..63a3df02 --- /dev/null +++ b/wiki/sources/nodewarden-ba-bitwarden-ban-dao-cloudflare-workers-shang-che-di-gao-bie-fu-wu-qi.md @@ -0,0 +1,39 @@ +--- +title: "NodeWarden - 把 Bitwarden 搬上 Cloudflare Workers,彻底告别服务器" +type: source +tags: [bitwarden, cloudflare, self-hosted, serverless] +date: 2026-02-22 +source_file: raw/Technical/Home Office/🟠NodeWarden - 把 Bitwarden 搬上 Cloudflare Workers,彻底告别服务器.md +--- + +## Summary +- 核心主题:NodeWarden——基于Cloudflare Workers的无服务器Bitwarden实现 +- 问题域:密码管理系统的极致轻量化部署 +- 方法/机制:Cloudflare D1存储密码 + R2存储附件 + Workers执行后端逻辑 +- 结论/价值:实现真正无服务器的密码管理,支持TOTP二次验证 + +## Key Claims +- NodeWarden将Bitwarden服务器端运行在Cloudflare Workers上,无需VPS +- 基于Cloudflare D1的单用户保管库和R2的附件存储实现完整Bitwarden兼容 +- 支持passkey和TOTP(通过TOTP_SECRET),官方Bitwarden需要付费会员 +- 定位单用户,放弃多用户、组织、集合等企业功能 + +## Key Quotes +> "部署 NodeWarden 之后的效果,就是在无服务器的情况下,也能在手机、电脑上使用 Bitwarden 客户端来保存密码了" — 核心价值主张 + +## Key Concepts +- [[无服务器密码管理]]:将传统服务器端逻辑迁移到边缘计算平台 +- [[Cloudflare D1]]:Cloudflare的SQLite数据库服务,用于存储密码元数据 +- [[Cloudflare R2]]:Cloudflare的对象存储服务,用于存储附件 + +## Key Entities +- [[Bitwarden]]:开源密码管理客户端,NodeWarden兼容的客户端 +- [[NodeWarden]]:运行在Cloudflare Workers上的Bitwarden服务器实现 +- [[Cloudflare]]:边缘计算平台,提供Workers/D1/R2服务 + +## Connections +- [[Cloudflare]] ← 部署平台 ← [[NodeWarden]] +- [[Bitwarden]] ← 客户端兼容 ← [[NodeWarden]] +- [[NodeWarden]] ← 替代方案 ← [[Bitwarden]] + +## Contradictions diff --git a/wiki/sources/notebooklm-open-source-alternatives.md b/wiki/sources/notebooklm-open-source-alternatives.md new file mode 100644 index 00000000..23a6f1f9 --- /dev/null +++ b/wiki/sources/notebooklm-open-source-alternatives.md @@ -0,0 +1,63 @@ +--- +title: "Google 神级生产力工具,所有 GitHub 开源平替都找到了" +type: source +tags: [notebooklm, open-source, ai-tools, github] +sources: ["https://mp.weixin.qq.com/s/6EoEMi8opDWOParUHRiHOg"] +last_updated: 2026-01-01 +--- + +## Summary + +- 核心主题:Google NotebookLM的开源替代方案汇总 +- 问题域:寻找可本地部署、功能类似的AI笔记和研究工具 +- 方法/机制:对比多个开源项目的功能特性、模型支持和部署方式 +- 结论/价值:为需要数据隐私控制或本地化需求的用户提供开源替代选择 + +## Key Claims + +- Open Notebook是GitHub上Star最高的NotebookLM开源替代(14.6k Stars),支持本地化知识管理和16+AI提供商 +- SurfSense定位为NotebookLM、Perplexity和Glean的开源替代(11.4k Stars),支持语义+全文混合搜索 +- Podcastfy专注播客生成,支持多语言文本转语音对话 +- 各开源项目在模型支持、部署方式、功能专注点上有差异化定位 + +## Key Quotes + +> "Open Notebook是一个全功能的本地化解决方案,不依赖云端的情况下进行知识管理和研究" — Open Notebook特点 + +> "SurfSense采用语义搜索+全文搜索混合搜索技术,并结合重排序算法,确保快速精准找到答案" — SurfSense技术特点 + +## Key Concepts + +- [[NotebookLM]]:Google AI笔记助手,基于用户文档进行问答,支持播客生成 +- [[Open Notebook]]:最高Star的开源替代,支持16+AI提供商和本地模型 +- [[SurfSense]]:AI搜索与研究智能体,支持Notion、YouTube、GitHub集成 +- [[Podcastfy]]:专注播客生成的工具,文本转多语言音频对话 +- [[notebookllama]]:LlamaIndex官方的文档转播客开源项目 +- [[PageLM]]:将学习材料转化为互动式资源的教育平台 +- [[InsightsLM]]:低代码/无代码的NotebookLM替代,基于Supabase和N8N + +## Key Entities + +- [[Open Notebook]]:开源项目,14.6k Stars +- [[SurfSense]]:开源项目,11.4k Stars +- [[Podcastfy]]:开源播客生成工具 +- [[notebookllama]]:LlamaIndex官方项目 +- [[PageLM]]:教育科技开源项目 +- [[InsightsLM]]:基于N8N的开源工具 +- [[NotebookLM]]:Google AI笔记产品 + +## Connections + +- [[NotebookLM]] ← 开源替代 ← [[Open Notebook]] +- [[NotebookLM]] ← 开源替代 ← [[SurfSense]] +- [[NotebookLM]] ← 开源替代 ← [[InsightsLM]] +- [[NotebookLM播客功能]] ← 开源替代 ← [[Podcastfy]] +- [[NotebookLM播客功能]] ← 开源替代 ← [[notebookllama]] +- [[学习工具]] ← 替代方案 ← [[PageLM]] + +## Contradictions + +- 与NotebookLM官方产品冲突: + - 冲突点:功能和用户体验完整性 + - 当前观点:开源方案提供本地化和定制化优势 + - 对方观点:官方产品功能更完整、体验更流畅 diff --git a/wiki/sources/openai-chatgpt-ge-xing-hua-ding-yi.md b/wiki/sources/openai-chatgpt-ge-xing-hua-ding-yi.md new file mode 100644 index 00000000..9ec4775f --- /dev/null +++ b/wiki/sources/openai-chatgpt-ge-xing-hua-ding-yi.md @@ -0,0 +1,28 @@ +--- +title: "OpenAI ChatGPT 个性化定义" +type: source +tags: [chatgpt, openai, 个性化, 自定义指令] +date: 2026-04-14 +--- + +## Summary +- 核心主题:ChatGPT自定义指令配置示例 +- 问题域:如何通过自定义指令让AI更符合个人需求 +- 方法/机制:提供个人背景信息+行为准则让AI精准适配 +- 结论/价值:展示一套完整的ChatGPT个性化配置模板 + +## Key Claims +- 自定义指令包含两部分:行为准则(Custom Instructions)和个人详情(Your Details) +- 行为准则:专业、详细、主动预判、不道德说教、引用来源 +- 个人详情:47岁自由职业者,前云服务交付高级经理,现专注TikTok跨境电商 + +## Key Concepts +- [[自定义指令]]:ChatGPT个性化配置 +- [[行为准则]]:对AI回复方式的偏好设定 +- [[个人详情]]:AI需要了解的用户背景信息 + +## Connections +- [[ChatGPT]] ← 配置 ← [[自定义指令]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/opencode-ubuntu-install.md b/wiki/sources/opencode-ubuntu-install.md new file mode 100644 index 00000000..4ac9cbef --- /dev/null +++ b/wiki/sources/opencode-ubuntu-install.md @@ -0,0 +1,38 @@ +--- +id: opencode-ubuntu-install +title: "在 Ubuntu 上安装 OpenCode 并配置 Vibe-Kanban" +type: source +tags: [opencode, ubuntu, vibe-coding, vibe-kanban] +sources: [] +last_updated: 2026-04-14 +--- + +## Summary +- 核心主题:OpenCode AI 编码 Agent 的安装配置及其与 Vibe-Kanban 的集成 +- 问题域:开源 AI 编码工具的本地部署和 LLM Provider 配置 +- 方法/机制:通过官方安装脚本一键安装,配置 API Key 后初始化项目 +- 结论/价值:OpenCode 支持任意 LLM Provider,通过 AGENTS.md 文件理解项目结构 + +## Key Claims +- OpenCode 可通过 `curl -fsSL https://opencode.ai/install | bash` 一键安装 +- OpenCode 支持任意 LLM Provider(需配置 API Key),也提供 OpenCode Zen 托管服务 +- 初始化项目后会自动创建 `AGENTS.md` 文件,建议提交到 Git +- OpenCode 提供 Plan 模式和 Build 模式切换(Tab 键),支持 /undo 和 /redo 命令 +- Vibe-Kanban 使用 GitHub CLI 创建 Pull Request + +## Key Concepts +- [[OpenCode]]:开源 AI 编码 Agent,支持终端界面、桌面应用和 IDE 扩展 +- [[Vibe-Kanban]]:任务看板,用于管理 AI 编码 Agent 的工作流 +- [[AGENTS-md]]:OpenCode 项目初始化时生成的文件,记录项目结构和编码规范 +- [[Plan模式]]:OpenCode 的规划模式,只生成实现计划不执行代码修改 +- [[Build模式]]:OpenCode 的构建模式,执行代码修改 + +## Key Entities +- [[BloopAI]]:OpenCode 的开发公司 + +## Connections +- [[OpenCode]] ← 支持 ← [[LLM-Provider]] +- [[Vibe-Kanban]] ← 集成 ← [[OpenCode]] +- [[OpenCode]] ← 生成 ← [[AGENTS-md]] + +## Contradictions diff --git a/wiki/sources/overnight-mini-app-builder.md b/wiki/sources/overnight-mini-app-builder.md new file mode 100644 index 00000000..977d5cb9 --- /dev/null +++ b/wiki/sources/overnight-mini-app-builder.md @@ -0,0 +1,31 @@ +--- +title: "Goal-Driven Autonomous Tasks" +type: source +tags: [OpenClaw, Autonomous, Goal-Driven, Kanban, Overnight] +sources: ["https://www.youtube.com/watch?v=41_TNGDDnfQ"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:目标驱动的自主任务 +- 问题域:如何让 AI Agent 主动生成并完成任务 +- 方法/机制:一次性灌输目标,Agent 每日自主生成/执行/追踪任务 +- 结论/价值:AI 作为自我导向的员工 + +## Key Claims +- 一次性灌输所有目标(职业/个人/商业) +- 每日生成 4-5 个自主完成的任务 +- 可构建惊喜夜间 mini-app(MVP) +- 竞态条件修复:分离 AUTONOMOUS.md(仅主会话写)+ tasks-log.md(仅追加) + +## Key Concepts +- [[Goal-Driven Autonomous]]:目标驱动自主 +- [[Overnight App Building]]:夜间应用构建 +- [[Race Condition Fix]]:竞态条件修复(双文件模式) + +## Connections +- [[OpenClaw]] ← 运行 ← [[Goal-Driven Autonomous]] +- [[Subagent]] ← 执行 ← [[Goal-Driven Autonomous]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/personal-crm.md b/wiki/sources/personal-crm.md new file mode 100644 index 00000000..f07d7308 --- /dev/null +++ b/wiki/sources/personal-crm.md @@ -0,0 +1,30 @@ +--- +title: "Personal CRM with Automatic Contact Discovery" +type: source +tags: [OpenClaw, CRM, Gmail, Calendar, Telegram, Contact Discovery] +date: 2026-04-14 +--- + +## Summary +- 核心主题:自动联系人发现的个人 CRM +- 问题域:如何追踪联系人互动历史和待跟进事项 +- 方法/机制:每日扫描邮件/日历,自动更新联系人数据库,会前简报 +- 结论/价值:不错过任何待跟进的人 + +## Key Claims +- 每日 cron 扫描邮件和日历,发现新联系人并更新 +- 自然语言查询:谁需要跟进?我上次和谁谈过? +- 每日会前简报:研究外部参会人背景、过往互动、待跟进事项 + +## Key Concepts +- [[Personal CRM]]:个人 CRM +- [[Contact Discovery]]:自动联系人发现 +- [[Meeting Prep]]:会议准备简报 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Personal CRM]] +- [[Gmail]] ← 数据源 ← [[Personal CRM]] +- [[Google Calendar]] ← 数据源 ← [[Personal CRM]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/phone-based-personal-assistant.md b/wiki/sources/phone-based-personal-assistant.md new file mode 100644 index 00000000..f2b6f03e --- /dev/null +++ b/wiki/sources/phone-based-personal-assistant.md @@ -0,0 +1,31 @@ +--- +title: "Phone-Based Personal Assistant" +type: source +tags: [OpenClaw, Phone, Voice, ClawdTalk, Telnyx] +sources: ["https://clawdtalk.com"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:电话语音助手 +- 问题域:如何在任何手机上免提访问 AI 助手 +- 方法/机制:ClawdTalk 让 OpenClaw 接收和拨打电话 +- 结论/价值:任何手机都成为 AI 助手网关 + +## Key Claims +- 通过电话呼叫与 AI Agent 对话 +- 获取日历提醒、Jira 更新、网络搜索结果的语音播报 +- Telnyx 集成提供可靠的电话连接 +- SMS 支持即将推出 + +## Key Concepts +- [[Phone-Based Assistant]]:电话语音助手 +- [[ClawdTalk]]:电话集成服务 +- [[Voice Gateway]]:语音网关 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Phone-Based Assistant]] +- [[ClawdTalk]] ← 电话服务 ← [[Phone-Based Assistant]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/phone-call-notifications.md b/wiki/sources/phone-call-notifications.md new file mode 100644 index 00000000..6bd8a225 --- /dev/null +++ b/wiki/sources/phone-call-notifications.md @@ -0,0 +1,31 @@ +--- +title: "Phone Call Notifications" +type: source +tags: [OpenClaw, Phone Call, Notifications, clawr.ing, Telnyx] +sources: ["https://clawr.ing"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:电话通知渠道 +- 问题域:如何让 Agent 电话通知重要事项 +- 方法/机制:clawr.ing 托管电话服务,Agent 主动拨号,双向对话 +- 结论/价值:唯一能保证被注意的通知渠道 + +## Key Claims +- Agent 主动拨号给用户,非用户呼叫 Agent +- clawr.ing 托管服务,无需 Twilio 配置 +- 覆盖 100+ 国家,PSTN 真实电话(非 VoIP) +- 不存储录音或转录,传输加密 + +## Key Concepts +- [[Phone Notification]]:电话通知渠道 +- [[clawr.ing]]:托管电话服务 +- [[Two-Way Conversation]]:双向对话 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Phone Call Notifications]] +- [[clawr.ing]] ← 电话服务 ← [[Phone Call Notifications]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/podcast-production-pipeline.md b/wiki/sources/podcast-production-pipeline.md new file mode 100644 index 00000000..7296e493 --- /dev/null +++ b/wiki/sources/podcast-production-pipeline.md @@ -0,0 +1,31 @@ +--- +title: "Podcast Production Pipeline" +type: source +tags: [OpenClaw, Podcast, Content Production, Show Notes, Social Media] +date: 2026-04-14 +--- + +## Summary +- 核心主题:播客生产流水线 +- 问题域:如何自动化播客制作 70% 的非录制工作 +- 方法/机制:研究 → 大纲/脚本 → 字幕处理 → 展示笔记 → 社媒工具包 +- 结论/价值:接题给题,收获完整生产包 + +## Key Claims +- 嘉宾/话题研究 + Talking Points 生成 +- 章节大纲 + 介绍脚本 + 过渡语生成 +- 字幕处理生成时间戳展示笔记 +- 社媒工具包:X/LinkedIn/Instagram 推广帖 +- SEO 优化的播客描述 + +## Key Concepts +- [[Podcast Production]]:播客生产 +- [[Content Pipeline]]:内容流水线 +- [[Show Notes]]:展示笔记 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Podcast Production Pipeline]] +- [[Content Factory]] ← 类似 ← [[Podcast Production Pipeline]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/polymarket-autopilot.md b/wiki/sources/polymarket-autopilot.md new file mode 100644 index 00000000..cfc76d7d --- /dev/null +++ b/wiki/sources/polymarket-autopilot.md @@ -0,0 +1,30 @@ +--- +title: "Polymarket Autopilot: Automated Paper Trading" +type: source +tags: [OpenClaw, Polymarket, Paper Trading, Discord, Prediction Market] +date: 2026-04-14 +--- + +## Summary +- 核心主题:Polymarket 自动Paper Trading +- 问题域:如何自动监控预测市场并执行交易策略 +- 方法/机制:监控市场数据,执行 TAIL/BONDING/SPREAD 策略,每日 Discord 报告 +- 结论/价值:用虚拟资金测试和优化交易策略 + +## Key Claims +- 监控 Polymarket API(价格/交易量/价差) +- TAIL(趋势跟随)、BONDING(逆向)、SPREAD(套利)三种策略 +- 每日 Discord 摘要:交易记录/P&L/胜率/策略表现 +- 从模式中学习:基于回测调整策略参数 + +## Key Concepts +- [[Polymarket Autopilot]]:预测市场自动交易 +- [[Paper Trading]]:虚拟资金交易测试 +- [[Trading Strategy]]:交易策略 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Polymarket Autopilot]] +- [[Discord]] ← 报告 ← [[Polymarket Autopilot]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/pre-build-idea-validator.md b/wiki/sources/pre-build-idea-validator.md new file mode 100644 index 00000000..bd58f83c --- /dev/null +++ b/wiki/sources/pre-build-idea-validator.md @@ -0,0 +1,31 @@ +--- +title: "Pre-Build Idea Validator" +type: source +tags: [OpenClaw, Idea Validation, MCP, GitHub, Hacker News] +sources: ["https://github.com/mnemox-ai/idea-reality-mcp"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:构建前创意验证 +- 问题域:如何在写代码前检查创意是否已存在 +- 方法/机制:MCP Server 扫描 GitHub/HN/npm/PyPI/Product Hunt,返回 reality_signal 评分 +- 结论/价值:避免解决已解决的问题 + +## Key Claims +- 扫描 5 个真实数据源构建前验证 +- 返回 reality_signal 评分(0-100) +- 显示 top 竞争对手及 star 数 +- 高分建议转向方向 + +## Key Concepts +- [[Pre-Build Validator]]:构建前验证 +- [[reality_signal]]:现实信号评分 +- [[MCP Server]]:MCP 服务器 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Pre-Build Validator]] +- [[idea-reality-mcp]] ← 工具 ← [[Pre-Build Validator]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/project-state-management.md b/wiki/sources/project-state-management.md new file mode 100644 index 00000000..cef865ce --- /dev/null +++ b/wiki/sources/project-state-management.md @@ -0,0 +1,30 @@ +--- +title: "Project State Management System: Event-Driven Alternative to Kanban" +type: source +tags: [OpenClaw, Project State, Event-Driven, PostgreSQL, Discord] +date: 2026-04-14 +--- + +## Summary +- 核心主题:事件驱动的项目状态管理系统 +- 问题域:如何替代静态 Kanban 追踪项目动态 +- 方法/机制:事件溯源替代卡片拖拽,自然语言更新,Git 集成 +- 结论/价值:对话式追踪项目状态,保留完整上下文 + +## Key Claims +- 事件日志替代静态看板:完成/阻塞/决定/转向均有记录 +- Git 提交自动关联项目,提交即文档 +- 每日 standup 摘要自动生成 +- 自然语言查询项目状态 + +## Key Concepts +- [[Project State Management]]:项目状态管理 +- [[Event Sourcing]]:事件溯源模式 +- [[Kanban Alternative]]:看板替代方案 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Project State Management]] +- [[PostgreSQL]] ← 存储 ← [[Project State Management]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/public-private-hybrid-cloud.md b/wiki/sources/public-private-hybrid-cloud.md new file mode 100644 index 00000000..3c6d2881 --- /dev/null +++ b/wiki/sources/public-private-hybrid-cloud.md @@ -0,0 +1,36 @@ +--- +title: "Public vs Private vs Hybrid Cloud Differences Explained" +type: source +tags: [Cloud, Public Cloud, Private Cloud, Hybrid Cloud] +sources: [] +last_updated: 2025-06-18 +--- + +## Summary +- 核心主题:公有云、私有云和混合云的差异与选择 +- 问题域:组织如何根据业务需求选择合适的云部署模式 +- 方法/机制:从安全性、成本、可扩展性、合规性等维度对比三种云模式 +- 结论/价值:平衡是云架构设计的关键驱动力,组织需根据工作负载特性选择最优模式 + +## Key Claims +- 公有云特征:高弹性、可扩展性、低成本订阅定价、快速运营化、最新技术、可靠性 +- 私有云特征:更高安全性、定制化控制、合规性、专用资源、无竞争延迟 +- 混合云优势:策略驱动部署、规模与安全性兼得、可靠性、成本控制与效率、互操作性和移动性、业务连续性 +- 共享责任模型:无论哪种云环境,组织都需对访问控制、云安全和灾难恢复规划负责 + +## Key Concepts +- [[公有云]]:通过互联网交付,多个组织共享的云服务模式 +- [[私有云]]:仅供单个组织使用的专用云环境 +- [[混合云]]:同时使用公有云和私有云的部署模式 +- [[共享责任模型]]:云提供商和组织共同承担安全和合规责任的框架 + +## Key Entities +- [[BMC]]:云咨询公司,发布云部署模式对比报告 + +## Connections +- [[公有云]] ← 适用 ← 非敏感工作负载 +- [[私有云]] ← 适用 ← 高合规要求场景 +- [[混合云]] ← 结合 ← 公有云与私有云优势 + +## Contradictions + diff --git a/wiki/sources/quan-wang-zui-quan-nano-banana-2-shi-yong-zhi-nan.md b/wiki/sources/quan-wang-zui-quan-nano-banana-2-shi-yong-zhi-nan.md new file mode 100644 index 00000000..117eb6cc --- /dev/null +++ b/wiki/sources/quan-wang-zui-quan-nano-banana-2-shi-yong-zhi-nan.md @@ -0,0 +1,38 @@ +--- +title: 全网最全!Nano Banana 2 使用指南(2025年12月更新) +type: source +tags: [Nano Banana 2, AI绘图, DeepSider] +date: 2025-12-01 +source_file: ../raw/AI/全网最全!Nano Banana 2 使用指南(2025年12月更新) 1.md +--- + +## Summary +- 核心主题:Nano Banana 2(Gemini 3 Pro Image)使用方法 +- 问题域:国内用户如何访问和使用Nano Banana 2 +- 方法/机制:通过DeepSider浏览器插件访问,支持GPT-5.1、Gemini 3.0等多种模型 +- 结论/价值:DeepSider是最方便的国内访问渠道之一,无需特殊网络 + +## Key Claims +- Nano Banana 2是推理模型,生成图像前会进行内部推理 +- 支持1K、2K、4K分辨率图像输出 +- 最多可将14张输入图像组合为1张输出图像 +- 擅长高事实准确性的创意工作、需要最新知识支持的图像创作 +- DeepSider是浏览器插件,安装后无需海外账户即可访问 +- 支持GPT-5.1、Gemini 2.5 Pro、Sora 2等热门模型 + +## Key Concepts +- [[Nano Banana 2]]:Google推理图像模型,Gemini 3 Pro Image代号 +- [[DeepSider]]:浏览器插件形式的AI模型聚合平台 +- [[推理模型]]:在生成图像前会进行内部推理的模型 + +## Key Entities +- [[DeepSider]]:浏览器插件,可访问Nano Banana 2 +- [[Nano Banana 2]]:Google AI图像生成模型 + +## Connections +- [[Nano Banana 2]] ← accessed_via ← [[DeepSider]] +- [[DeepSider]] ← includes ← [[Gemini 3 Pro]] +- [[DeepSider]] ← includes ← [[GPT-5.1]] +- [[DeepSider]] ← includes ← [[Sora 2]] + +## Contradictions diff --git a/wiki/sources/qun-hui-nas-ke-xue-shang-wang-fang-fa.md b/wiki/sources/qun-hui-nas-ke-xue-shang-wang-fang-fa.md new file mode 100644 index 00000000..222ee4f9 --- /dev/null +++ b/wiki/sources/qun-hui-nas-ke-xue-shang-wang-fang-fa.md @@ -0,0 +1,32 @@ +--- +title: "群晖NAS科学上网方法" +type: source +tags: [Synology, NAS, V2RayA, Docker] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠群晖NAS科学上网方法.md +--- + +## Summary +- 核心主题:群晖NAS通过V2RayA配置透明代理 +- 问题域:NAS Docker pull速度慢/被墙 +- 结论/价值:V2RayA透明代理+Docker Daemon HTTP代理配置 + +## Key Claims +- V2RayA Docker镜像安装,Web界面配置 +- 透明代理对Docker Daemon可能无效 +- 最终方案:配置Docker Daemon的HTTP_PROXY环境变量 +- 显式配置代理比透明代理更符合Engineering Best Practice + +## Key Concepts +- [[V2RayA]]:v2rayA Docker透明代理 +- [[Docker代理配置]]:systemd配置HTTP_PROXY +- [[透明代理]]:Host模式透明劫持流量 + +## Key Entities +- [[Synology-NAS]]:群晖NAS设备 + +## Connections +- [[RAX50梅林固件]] ← 替代方案 ← [[群晖NAS科学上网]] + +## Contradictions +- 与透明代理优先思维冲突:本文认为显式配置Docker Daemon代理更稳定 diff --git a/wiki/sources/rag-cong-ru-men-dao-jing-tong-xi-lie-1.md b/wiki/sources/rag-cong-ru-men-dao-jing-tong-xi-lie-1.md new file mode 100644 index 00000000..785620eb --- /dev/null +++ b/wiki/sources/rag-cong-ru-men-dao-jing-tong-xi-lie-1.md @@ -0,0 +1,41 @@ +--- +title: "RAG从入门到精通系列1:基础RAG" +type: source +tags: [RAG, LLM, 检索增强生成, 向量数据库] +sources: ["https://mp.weixin.qq.com/s/TlFNOw7_3Q8qywKLpVUmfg"] +date: 2025-12-18 +--- + +## Summary +- 核心主题:RAG基础概念与实现流程 +- 问题域:如何将LLM与外部数据源连接 +- 方法/机制:Indexing → Retrieval → Generation三阶段流程 +- 结论/价值:RAG是连接LLM与外部私有/最新数据的通用方法 + +## Key Claims +- RAG = Retrieval Augmented Generation,将LLM与外部数据源连接 +- 三阶段:Indexing(文档索引)、Retrieval(检索)、Generation(生成) +- Embedding Model将文本转为固定长度向量,受Context Window限制 +- 向量数据库(Qdrant等)存储Embedding Vector并实现相似度检索 +- LangChain/LlamaIndex等框架简化RAG流程 + +## Key Concepts +- [[RAG]](检索增强生成):将LLM与外部数据源连接的通用方法 +- [[Indexing]]:将外部文档切分并建立索引的过程 +- [[Retrieval]]:根据问题语义向量检索相关文档 +- [[Generation]]:将问题+知识片段输入LLM生成答案 +- [[Embedding Vector]]:文本的数值表示,捕获语义信息 +- [[Vector Store]]:存储Embedding Vector并实现相似度检索的数据库 +- [[Context Window]]:Embedding Model能接受的最大token数 + +## Connections +- [[RAG]] ← 包含 ← [[Indexing]] +- [[RAG]] ← 包含 ← [[Retrieval]] +- [[RAG]] ← 包含 ← [[Generation]] +- [[LLM]] ← 扩展 ← [[RAG]] + +## Contradictions +- 与[[LLM Wiki]]冲突: + - 冲突点:知识获取方式 + - 当前观点:RAG从外部检索获取最新/私有数据 + - [[LLM Wiki]]观点:增量累积知识到内部记忆 diff --git a/wiki/sources/rax50-lu-you-qi-geng-xin-merlin-clash-ding-yue.md b/wiki/sources/rax50-lu-you-qi-geng-xin-merlin-clash-ding-yue.md new file mode 100644 index 00000000..1e1943b7 --- /dev/null +++ b/wiki/sources/rax50-lu-you-qi-geng-xin-merlin-clash-ding-yue.md @@ -0,0 +1,30 @@ +--- +title: "RAX50 路由器 更新Merlin Clash订阅" +type: source +tags: [clash, merlin-clash, rax50, router] +date: 2026-04-13 +source_file: raw/Technical/Home Office/🟠RAX50 路由器 更新Merlin Clash订阅.md +--- + +## Summary +- 核心主题:Netgear RAX50路由器刷Merlin固件后更新Clash订阅配置 +- 问题域:家庭网络科学上网配置 +- 方法/机制:在Merlin Clash界面通过小白一键订阅助手导入VLESS URL +- 结论/价值:完成路由器端科学上网配置,实现局域网透明代理 + +## Key Claims +- 进入RAX50路由器管理界面的Merlin Clash页面 +- 通过小白一键订阅助手导入VLESS URL并重命名配置文件 +- 选择新建配置文件后点保存与启动,如不生效需再次快速重启 + +## Key Concepts +- [[路由器科学上网]]:在路由器层面配置代理,实现局域网设备透明代理 + +## Key Entities +- [[Merlin Clash]]:基于Clash的梅林固件科学上网插件 +- [[RAX50]]:Netgear RAX50路由器型号 + +## Connections +- [[Merlin Clash]] ← 运行于 ← [[RAX50]] + +## Contradictions diff --git a/wiki/sources/rto-vs-rpo-disaster-recovery.md b/wiki/sources/rto-vs-rpo-disaster-recovery.md new file mode 100644 index 00000000..7965671a --- /dev/null +++ b/wiki/sources/rto-vs-rpo-disaster-recovery.md @@ -0,0 +1,39 @@ +--- +title: "RTO vs RPO Key Differences for Modern Disaster Recovery" +type: source +tags: [RTO, RPO, Disaster Recovery, Feature Flags] +sources: [] +last_updated: 2025-07-26 +--- + +## Summary +- 核心主题:RTO与RPO的区别及其在现代灾难恢复中的应用 +- 问题域:团队如何为现代软件交付设置切实可行的恢复目标 +- 方法/机制:通过功能标志实现秒级RTO和低RPO +- 结论/价值:预防优于恢复,将韧性直接嵌入开发工作流 + +## Key Claims +- RTO(恢复时间目标):系统可容忍的最大停机时间 +- RPO(恢复点目标):可接受的最大数据丢失量 +- 每日部署场景下,每次部署都是潜在的RTO/RPO场景 +- 现代主要宕机来源是代码变更(支付bug、数据库迁移、AI模型更新)而非硬件故障 +- 功能标志可将回滚时间从数小时缩短到秒级 +- 渐进式发布(1%→5%→25%→100%)可将RTO控制在秒级 + +## Key Concepts +- [[RTO]]:Recovery Time Objective,恢复时间目标 +- [[RPO]]:Recovery Point Objective,恢复点目标 +- [[功能标志]]:控制功能开启/关闭的开关,无需重新部署 +- [[渐进式发布]]:逐步向用户推出新功能的发布策略 +- [[Kill Switch]]:紧急情况下立即关闭功能的开关 + +## Key Entities +- [[LaunchDarkly]]:功能管理平台 + +## Connections +- [[RTO]] ← 关联 ← [[功能标志]] +- [[RPO]] ← 关联 ← 数据完整性 +- [[功能标志]] ← 实现 ← [[Kill Switch]] + +## Contradictions + diff --git a/wiki/sources/ru-he-chuan-shu-docker-images-bing-qie-zai-yi-ge-docker-an-zhuang.md b/wiki/sources/ru-he-chuan-shu-docker-images-bing-qie-zai-yi-ge-docker-an-zhuang.md new file mode 100644 index 00000000..c43df915 --- /dev/null +++ b/wiki/sources/ru-he-chuan-shu-docker-images-bing-qie-zai-yi-ge-docker-an-zhuang.md @@ -0,0 +1,33 @@ +--- +title: "如何传输Docker images并且在另一个Docker安装" +type: source +tags: [docker, synology, nas] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠如何传输Docker images 并且在另一个Docker安装.md +--- + +## Summary +- 核心主题:将Docker镜像从一台设备传输到另一台设备 +- 问题域:Docker镜像迁移与离线安装 +- 方法/机制:使用docker save导出为tar文件,传输后用docker load导入 +- 结论/价值:实现跨设备的Docker镜像迁移,无需重新pull + +## Key Claims +- 使用docker pull从Docker Hub拉取镜像 +- 使用docker save将镜像打包为tar文件用于传输 +- 使用docker load在目标设备导入镜像 + +## Key Concepts +- [[Docker镜像导出]]:docker save将镜像保存为tar归档 +- [[Docker镜像导入]]:docker load从tar归档恢复镜像 + +## Key Entities +- [[Docker]]:容器化平台 +- [[Synology NAS]]:带有Docker套件的NAS设备 +- [[Docker Hub]]:Docker镜像仓库 + +## Connections +- [[Docker]] ← save导出 ← [[Docker镜像导出]] +- [[Docker镜像导入]] ← load导入 ← [[Docker]] + +## Contradictions diff --git a/wiki/sources/ru-he-li-yong-sora-jie-kou-shi-xian-shi-pin-zi-dong-hua-sheng-cheng-gong-zuo-liu.md b/wiki/sources/ru-he-li-yong-sora-jie-kou-shi-xian-shi-pin-zi-dong-hua-sheng-cheng-gong-zuo-liu.md new file mode 100644 index 00000000..c4eea541 --- /dev/null +++ b/wiki/sources/ru-he-li-yong-sora-jie-kou-shi-xian-shi-pin-zi-dong-hua-sheng-cheng-gong-zuo-liu.md @@ -0,0 +1,37 @@ +--- +title: "如何利用Sora接口实现视频自动化生成工作流" +type: source +tags: [sora, n8n, workflow, video-generation] +date: 2025-12-18 +source_file: raw/AI/如何利用Sora接口实现视频自动化生成工作流.md +--- + +## Summary +- 核心主题:Sora接口实现视频生成全自动化工作流 +- 问题域:AI视频生成、自媒体内容创作 +- 方法/机制:Sora API + n8n工作流自动化 +- 结论/价值:低成本高效率的视频生成方案,适合自媒体副业 + +## Key Claims +- Sora接口成本比OpenAI便宜六倍以上 +- 新用户注册亚马逊账户可享受200美元抵扣金和六个月免费试用 +- 使用Sora生成一般视频费用仅需两三元人民币 +- 精细化的提示词设计能够显著提升生成视频的质量 + +## Key Quotes +> "通过Sora接口实现视频生成的经济实惠方案" — 核心价值主张 +> "提示词的艺术:提高生成内容质量的关键在于精细化的提示词设计" — 欧阳 + +## Key Concepts +- [[视频自动化生成工作流]]:通过API接口实现视频生成的批量自动化 +- [[UGC创作]]:用户生成内容,适用于市场推广 +- [[提示词优化]]:提升AI生成内容质量的关键技术 + +## Key Entities +- [[Sora]]:OpenAI的视频生成模型 +- [[n8n]]:工作流自动化工具 + +## Connections +- [[n8n]] ← 工作流编排 ← [[Sora]] + +## Contradictions diff --git a/wiki/sources/ru-he-pan-bie-ni-de-linux-fu-wu-qi-shi-x64-huan-shi-arm64.md b/wiki/sources/ru-he-pan-bie-ni-de-linux-fu-wu-qi-shi-x64-huan-shi-arm64.md new file mode 100644 index 00000000..ca87d549 --- /dev/null +++ b/wiki/sources/ru-he-pan-bie-ni-de-linux-fu-wu-qi-shi-x64-huan-shi-arm64.md @@ -0,0 +1,33 @@ +--- +title: "如何判别你的Linux服务器是x64还是ARM64" +type: source +tags: [linux] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠如何判别你的Linux 服务器是 x64(也就是 x86_64)还是 ARM64.md +--- + +## Summary +- 核心主题:识别Linux服务器的CPU架构类型 +- 问题域:系统架构检测 +- 方法/机制:使用uname、lscpu、/proc/cpuinfo、file命令 +- 结论/价值:确认架构以便安装对应版本的软件 + +## Key Claims +- uname -m输出x86_64表示Intel/AMD 64位架构,aarch64表示ARM64 +- lscpu的Architecture字段直接显示CPU类型 +- /proc/cpuinfo中model name显示Intel类型,ARM64显示AArch64/ARMv8 +- file命令检测可执行文件架构 + +## Key Concepts +- [[x86_64]]:Intel/AMD 64位处理器架构 +- [[ARM64]]:64位ARM处理器架构,又称aarch64 +- [[CPU架构检测]]:通过命令行识别系统架构 + +## Key Entities +- [[Linux]]:开源操作系统内核 + +## Connections +- [[Linux]] ← 运行在 ← [[x86_64]] +- [[Linux]] ← 运行在 ← [[ARM64]] + +## Contradictions diff --git a/wiki/sources/ru-he-shan-chu-jiu-de-fei-qi-de-docker-container-volume.md b/wiki/sources/ru-he-shan-chu-jiu-de-fei-qi-de-docker-container-volume.md new file mode 100644 index 00000000..1ee1d8b1 --- /dev/null +++ b/wiki/sources/ru-he-shan-chu-jiu-de-fei-qi-de-docker-container-volume.md @@ -0,0 +1,35 @@ +--- +title: "如何删除旧的废弃的docker container + volume" +type: source +tags: [docker, container, volume] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠如何删除旧的废弃的docker container +volume.md +--- + +## Summary +- 核心主题:清理旧的Docker容器、卷和网络 +- 问题域:Docker资源清理 +- 方法/机制:使用docker命令停止、删除容器和卷 +- 结论/价值:解决WARN提示(网络/卷已存在但不属于当前项目) + +## Key Claims +- docker rm -f强制删除运行中的容器 +- docker volume rm删除悬空卷 +- docker network rm删除孤立网络 +- docker compose down可一次性清理整个compose项目 + +## Key Concepts +- [[Docker容器清理]]:停止并删除旧容器 +- [[Docker卷管理]]:清理未使用的Docker数据卷 +- [[Docker网络清理]]:删除孤立网络 + +## Key Entities +- [[Docker]]:容器化平台 +- [[Portainer]]:Docker可视化管理系统 + +## Connections +- [[Docker]] ← 管理 ← [[Docker容器清理]] +- [[Docker]] ← 管理 ← [[Docker卷管理]] +- [[Docker]] ← 管理 ← [[Docker网络清理]] + +## Contradictions diff --git a/wiki/sources/ru-he-xie-chu-wan-mei-de-prompt.md b/wiki/sources/ru-he-xie-chu-wan-mei-de-prompt.md new file mode 100644 index 00000000..9d351139 --- /dev/null +++ b/wiki/sources/ru-he-xie-chu-wan-mei-de-prompt.md @@ -0,0 +1,47 @@ +--- +title: 如何写出完美的Prompt(提示词)? +type: source +tags: [Prompt, AI提示词, 提示词工程] +date: 2025-12-02 +source_file: ../raw/AI/如何写出完美的Prompt(提示词)?.md +--- + +## Summary +- 核心主题:构建高效Prompt的底层逻辑和实操方法 +- 问题域:如何将模糊需求转化为AI可理解的结构化任务 +- 方法/机制:误区分析 → 底层逻辑 → 分层实操(基础/进阶/高阶) → 场景模板 +- 结论/价值:Prompt能力本质是需求拆解能力+结构化表达能力+场景共情能力+迭代优化能力 + +## Key Claims +- Prompt不仅是给AI的指令,更是一套人与AI的协作协议 +- Prompt能力本质:清晰界定核心诉求+建立与AI能力匹配的沟通逻辑 +- 三大误区:过度堆砌术语、忽视隐性需求、期望一键生成 +- 底层逻辑:角色、需求、场景、目标缺一不可 +- 基础方法:需求拆解法、上下文补全法、格式定义法、示例引导法 +- 进阶策略:思维链引导法、任务拆分法、角色赋能法、预填回复法、不确定性管理法 +- 高阶技巧:跨模态联动法、领域知识注入法、反馈循环嵌入法 +- 五大坑:过度设计、基础缺失、不明确隐性需求、技巧堆砌、忽视伦理 + +## Key Concepts +- [[Prompt能力]]:清晰界定核心诉求+建立与AI能力匹配的沟通逻辑 +- [[需求拆解]]:将模糊目标转化为具体、可执行的子任务 +- [[结构化表达]]:用清晰的逻辑组织信息 +- [[上下文补全]]:提供AI所需的关键背景信息 +- [[示例引导]]:用少量样本提示解决风格/格式难题 +- [[思维链]]:让AI逐步推理的引导方法 +- [[任务拆分]]:将复杂任务分解为多个子任务 +- [[角色赋能]]:让AI代入专业视角 +- [[预填回复]]:强制输出结构化格式 +- [[不确定性管理]]:明确告知AI不知道就标注 + +## Key Entities +- [[Claude]]:Anthropic大模型 +- [[GPT-4]]:OpenAI大模型 + +## Connections +- [[Prompt能力]] ← requires ← [[需求拆解]] +- [[Prompt能力]] ← requires ← [[结构化表达]] +- [[Prompt能力]] ← requires ← [[场景共情能力]] +- [[Prompt能力]] ← requires ← [[迭代优化能力]] + +## Contradictions diff --git a/wiki/sources/ru-he-yong-zhi-wen-liu-lan-qi-an-quan-zhu-ce-bing-ding-yue-claude-pro-hui-yuan-quan-gong-lue.md b/wiki/sources/ru-he-yong-zhi-wen-liu-lan-qi-an-quan-zhu-ce-bing-ding-yue-claude-pro-hui-yuan-quan-gong-lue.md new file mode 100644 index 00000000..caf82b35 --- /dev/null +++ b/wiki/sources/ru-he-yong-zhi-wen-liu-lan-qi-an-quan-zhu-ce-bing-ding-yue-claude-pro-hui-yuan-quan-gong-lue.md @@ -0,0 +1,65 @@ +--- +title: "如何用指纹浏览器安全注册并订阅Claude Pro会员全攻略" +type: source +tags: [claude, ads-power, pingme, wildcard, fingerprint-browser, 指纹浏览器] +date: 2025-12-31 +source_file: raw/Technical/Home Office/🟠如何用指纹浏览器安全注册并订阅Claude Pro会员全攻略.md +--- + +## Summary +- 核心主题:指纹浏览器 + 代理IP + 虚拟信用卡注册Claude Pro +- 问题域:账号封禁、跨境支付、验证码接收 +- 方法/机制:AdsPower隔离环境 + 美国代理 + PingMe接码 + WildCard支付 +- 结论/价值:安全注册并订阅Claude Pro的完整方案 + +## Key Claims +- 指纹浏览器隔离环境防止账号关联 +- IP纯净度检测(scamalytics低风险)降低封号风险 +- PingMe订阅制接码平台提供稳定验证码 +- WildCard虚拟信用卡解决跨境支付问题 + +## Key Concepts +- [[指纹浏览器]]:模拟独立设备环境隔离账号 +- [[SOCKS5代理]]:网络代理协议,隐匿真实IP +- [[IP纯净度]]:IP信誉风险等级,低风险才安全 +- [[虚拟信用卡]]:WildCard等支持海外支付的线上信用卡 + +## 工具推荐 +| 工具 | 用途 | +|------|------| +| AdsPower | 指纹浏览器,隔离环境 | +| SOCKS5代理 | 美国高质量代理IP | +| PingMe | 短信接码平台(订阅制) | +| WildCard | 虚拟信用卡(支持海外支付) | + +## 操作流程 + +### 1. 环境配置 +- 创建新浏览器环境 +- 选择Chrome最新版本 + Windows操作系统 +- 配置SOCKS5代理(从系统代理复制地址端口) + +### 2. IP验证 +- 访问ip111.cn确认IP为美国 +- 多网站验证IP一致性(国内外+谷歌三处一致) +- 用scamalytics.com检测IP纯净度(需低风险) + +### 3. Claude注册 +- 用谷歌账号登录Claude +- 通过PingMe接收验证码(美国区号码) +- 推荐号码:+1 914 577 5122 + +### 4. Claude Pro订阅 +- 注册WildCard虚拟信用卡 +- 支付宝充值(最低20美元) +- 绑定信用卡完成订阅 + +## 关键点 +- IP必须一致:代理设置的IP与检测网站显示的IP必须匹配 +- 纯净度必须低:scamalytics风险值越低越安全 +- 接码用订阅制:避免一次性号码被封 +- 支付用虚拟卡:国内信用卡无法支付海外服务 + +## 账户信息 +- Google登录:billyshen2000@gmail.com +- Claude账号:通过谷歌登录 \ No newline at end of file diff --git a/wiki/sources/ru-he-zai-ubuntu-server-an-zhuang-docker-compose.md b/wiki/sources/ru-he-zai-ubuntu-server-an-zhuang-docker-compose.md new file mode 100644 index 00000000..113590bf --- /dev/null +++ b/wiki/sources/ru-he-zai-ubuntu-server-an-zhuang-docker-compose.md @@ -0,0 +1,34 @@ +--- +title: "如何在Ubuntu Server安装docker & docker compose" +type: source +tags: [docker, ubuntu] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠如何在Ubuntu Server安装 docker & docker compose.md +--- + +## Summary +- 核心主题:在Ubuntu Server上安装Docker Engine和Docker Compose V2 +- 问题域:Docker环境搭建 +- 方法/机制:从Docker官方仓库安装,使用apt包管理器 +- 结论/价值:完整安装Docker环境,支持非root用户运行 + +## Key Claims +- 应从Docker官方仓库安装以获取最新版本 +- docker-compose-plugin安装的是Docker Compose V2,使用docker compose命令 +- 将用户加入docker组后可无sudo运行Docker命令 +- 安装后需要重新登录或执行newgrp docker使组权限生效 + +## Key Concepts +- [[Docker Engine]]:Docker核心运行时 +- [[Docker Compose V2]]:Docker多容器编排工具 +- [[Docker官方仓库]]:Docker的apt软件源 + +## Key Entities +- [[Docker]]:容器化平台 +- [[Ubuntu]]:Linux发行版 + +## Connections +- [[Ubuntu]] ← 安装 ← [[Docker Engine]] +- [[Docker Engine]] ← 包含 ← [[Docker Compose V2]] + +## Contradictions diff --git a/wiki/sources/ru-he-zai-ubuntu-server-shang-tong-guo-nfs-gua-zai-synology-nas-shang-de-gong-xiang-wen-jian-jia.md b/wiki/sources/ru-he-zai-ubuntu-server-shang-tong-guo-nfs-gua-zai-synology-nas-shang-de-gong-xiang-wen-jian-jia.md new file mode 100644 index 00000000..dbca4dae --- /dev/null +++ b/wiki/sources/ru-he-zai-ubuntu-server-shang-tong-guo-nfs-gua-zai-synology-nas-shang-de-gong-xiang-wen-jian-jia.md @@ -0,0 +1,35 @@ +--- +title: "如何在Ubuntu Server上通过NFS挂载Synology NAS上的共享文件夹" +type: source +tags: [nfs, synology, ubuntu] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠如何在Ubuntu Server上通过NFS挂载Synology NAS上的共享文件夹.md +--- + +## Summary +- 核心主题:通过NFS协议将Synology NAS共享文件夹挂载到Ubuntu Server +- 问题域:NAS与Linux服务器之间的网络存储挂载 +- 方法/机制:NAS端配置NFS权限,Ubuntu端使用mount命令挂载,配置fstab实现永久挂载 +- 结论/价值:实现NAS作为备份存储的稳定挂载 + +## Key Claims +- NFS相比Samba能完美保留Linux文件所有权信息,避免Docker卷恢复时的权限报错 +- NFS挂载需要使用timeo=900和_netdev参数确保稳定性 +- 映射所有用户为admin可绕过复杂的Linux权限校验 +- 必须使用mount -a测试而非直接重启验证fstab配置 + +## Key Concepts +- [[NFS]]:Network File System,网络文件系统协议 +- [[fstab]]:Linux文件系统表,用于自动挂载配置 +- [[永久挂载]]:通过fstab实现开机自动挂载 + +## Key Entities +- [[Synology NAS]]:网络附加存储设备 +- [[Ubuntu]]:Linux发行版 + +## Connections +- [[Ubuntu]] ← 挂载 ← [[NFS]] +- [[NFS]] ← 服务于 ← [[Synology NAS]] +- [[fstab]] ← 配置 ← [[永久挂载]] + +## Contradictions diff --git a/wiki/sources/second-brain.md b/wiki/sources/second-brain.md new file mode 100644 index 00000000..2700522f --- /dev/null +++ b/wiki/sources/second-brain.md @@ -0,0 +1,31 @@ +--- +title: "Second Brain" +type: source +tags: [OpenClaw, Second Brain, Memory, Next.js, Search] +sources: ["https://www.youtube.com/watch?v=41_TNGDDnfQ"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:第二大脑 +- 问题域:如何让记忆捕获和检索像发短信一样简单 +- 方法/机制:Telegram/iMessage/Discord 文本捕获 + Next.js 可搜索仪表板 +- 结论/价值:零摩擦捕获 + 语义检索 + +## Key Claims +- 文本发送任何内容给 OpenClaw,即刻记住 +- OpenClaw 内置记忆系统永久存储 +- Next.js 仪表板全局搜索所有记忆/对话/笔记 +- 无文件夹/无标签/无复杂组织,只有文本和搜索 + +## Key Concepts +- [[Second Brain]]:第二大脑 +- [[Zero-Friction Capture]]:零摩擦捕获 +- [[Semantic Search]]:语义检索 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Second Brain]] +- [[Telegram]] ← 捕获渠道 ← [[Second Brain]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/self-healing-home-server.md b/wiki/sources/self-healing-home-server.md new file mode 100644 index 00000000..7a086274 --- /dev/null +++ b/wiki/sources/self-healing-home-server.md @@ -0,0 +1,32 @@ +--- +title: "Self-Healing Home Server & Infrastructure Management" +type: source +tags: [OpenClaw, Home Server, Self-Healing, SSH, Kubernetes] +sources: ["https://madebynathan.com/2026/02/03/everything-ive-done-with-openclaw-so-far/"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:自愈家庭服务器与基础设施管理 +- 问题域:如何让家庭服务器 24/7 自我监控和修复 +- 方法/机制:SSH 访问 + 自动 cron 作业 + 健康检查 + 自愈修复 +- 结论/价值:AI 会硬编码密钥,必须安全优先 + +## Key Claims +- 自动健康监控:服务/部署/系统资源 +- 自愈:健康检查检测问题,自主应用修复(重启 pod/扩展资源/修复配置) +- 早晨简报:天气/日历/系统状态/任务面板 +- 安全审计:定期扫描硬编码密钥/特权容器/过度授权访问 + +## Key Concepts +- [[Self-Healing Server]]:自愈服务器 +- [[Infrastructure Agent]]:基础设施 Agent +- [[Security Audit]]:安全审计 + +## Connections +- [[OpenClaw]] ← 运行 ← [[Self-Healing Home Server]] +- [[SSH]] ← 访问 ← [[Self-Healing Home Server]] +- [[Kubernetes]] ← 集群管理 ← [[Self-Healing Home Server]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/semantic-memory-search.md b/wiki/sources/semantic-memory-search.md new file mode 100644 index 00000000..79bc9e06 --- /dev/null +++ b/wiki/sources/semantic-memory-search.md @@ -0,0 +1,31 @@ +--- +title: "Semantic Memory Search" +type: source +tags: [OpenClaw, Semantic Search, memsearch, Milvus, Vector DB] +sources: ["https://github.com/zilliztech/memsearch"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:语义记忆搜索 +- 问题域:如何在 OpenClaw 增长的 markdown 记忆中语义检索 +- 方法/机制:memsearch + Milvus 向量数据库,混合搜索 + BM25 + RRF 重排 +- 结论/价值:SHA-256 内容哈希,零浪费 API 调用 + +## Key Claims +- 向量搜索 + BM25 全文搜索混合,RRF 重排 +- SHA-256 内容哈希,未更改文件永不重嵌入 +- 文件监视器自动重索引,索引始终最新 +- 支持 OpenAI/Google/Voyage/Ollama/完全本地嵌入 + +## Key Concepts +- [[Semantic Memory Search]]:语义记忆搜索 +- [[memsearch]]:向量搜索工具 +- [[Hybrid Search]]:混合搜索 + +## Connections +- [[OpenClaw]] ← 数据源 ← [[Semantic Memory Search]] +- [[Milvus]] ← 向量数据库 ← [[Semantic Memory Search]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/synology-nas-xiaoya-alist-clouddrive2-plex-media-platform.md b/wiki/sources/synology-nas-xiaoya-alist-clouddrive2-plex-media-platform.md new file mode 100644 index 00000000..a6b684bc --- /dev/null +++ b/wiki/sources/synology-nas-xiaoya-alist-clouddrive2-plex-media-platform.md @@ -0,0 +1,66 @@ +--- +title: "Synology NAS + Xiaoya Alist + CloudDrive2 + Plex 家庭影视平台" +type: source +tags: [synology, nas, xiaoya, alist, clouddrive2, plex, media] +date: 2025-02-23 +source_file: raw/Technical/Home Office/Synology NAS + Xiaoya Alist + CloudDrvie2+ Plex to Build Media Platform.md +--- + +## Summary +- 核心主题:构建以NAS为中心的家庭媒体平台 +- 问题域:云盘资源整合、媒体服务部署、多设备播放 +- 方法/机制:CloudDrive2挂载阿里云盘 + Xiaoya Alist资源整理 + Plex统一播放 +- 结论/价值:实现阿里云盘资源的本地化播放体验 + +## Key Claims +- CloudDrive2可挂载阿里云盘为本地文件系统 +- Xiaoya Alist提供资源聚合和转存功能 +- Plex支持多平台客户端播放 +- Docker方式部署确保隔离和便携性 + +## Key Concepts +- [[云盘挂载]]:CloudDrive2将云盘映射为本地 +- [[资源聚合]]:Xiaoya Alist整合多源资源 +- [[媒体服务器]]:Plex家庭媒体中心 +- [[Docker]]:容器化部署 + +## Key Entities +- [[Synology NAS]]:NAS设备,媒体平台核心 +- [[Plex]]:媒体服务器软件 +- [[Xiaoya Alist]]:资源聚合工具 +- [[CloudDrive2]]:云盘挂载工具 +- [[阿里云盘]]:云存储资源 + +## 部署步骤 + +### 1. Plex安装 +- 群晖套件中心直接安装Plex Media Server +- 客户端:华为盒子(Android 5.0+)、当贝盒子 + +### 2. Xiaoya Alist安装 +- 通过Docker加载xiaoya/alist镜像 +- 配置token文件:myopentoken.txt, mytoken.txt, temp_transfer_folder_id.txt +- 访问地址:http://192.168.3.17:5678/ + +### 3. CloudDrive2安装 +- 添加矿神源后安装 +- DSM 7+需执行:sudo sed -i 's/package/root/g' /var/packages/CloudDrive2/conf/privilege +- 访问地址:http://192.168.3.17:19798/ +- 仅授权资源目录,不授权备份目录 + +### 4. Plex媒体库配置 +- 通过Xiaoya选择资源,移到aliyun-movie、aliyun-tvshows等目录 +- Plex自动刮削并展示 + +## 目录结构 +``` +NAS +├── docker/xiaoya/ # Xiaoya配置 +│ ├── myopentoken.txt +│ ├── mytoken.txt +│ └── temp_transfer_folder_id.txt +├── Software/家庭影视平台/Plex # Plex APK文件 +└── aliyun-movie/ # 媒体目录 + aliyun-tvshows/ + aliyun-documentory/ +``` \ No newline at end of file diff --git a/wiki/sources/templater-plugin.md b/wiki/sources/templater-plugin.md new file mode 100644 index 00000000..372e002b --- /dev/null +++ b/wiki/sources/templater-plugin.md @@ -0,0 +1,31 @@ +--- +title: "Templater Obsidian Plugin" +type: source +tags: [Obsidian, Plugin, Template] +date: 2026-04-14 +source_file: raw/Technical/Templater Obsidian Plugin.md +--- + +## Summary +- 核心主题:Templater插件的功能介绍 +- 问题域:Obsidian模板语言,支持变量插入和JavaScript执行 +- 方法/机制:模板语法 + JavaScript脚本扩展 +- 结论/价值:高度灵活的Obsidian模板解决方案 + +## Key Claims +- Templater定义了一种模板语言,支持插入变量和函数结果 +- 支持执行JavaScript代码操作变量和函数 +- 警告:执行任意JavaScript和系统命令可能有安全风险 + +## Key Concepts +- [[Templater]]:Obsidian模板插件,支持变量和JavaScript + +## Key Entities +- [[Obsidian]]:笔记管理工具 +- [[SilentVoid13]]:Templater开发者 + +## Connections +- [[Obsidian]] ← 插件 ← [[Templater]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/the-picture-they-paint-of-you.md b/wiki/sources/the-picture-they-paint-of-you.md new file mode 100644 index 00000000..1150d0e1 --- /dev/null +++ b/wiki/sources/the-picture-they-paint-of-you.md @@ -0,0 +1,38 @@ +--- +title: "The Picture They Paint of You" +type: source +tags: [AI工具, 框架分析, SRE, 编码助手, 社会学] +sources: ["https://ferd.ca/the-picture-they-paint-of-you.html"] +date: 2026-04-13 +--- + +## Summary +- 核心主题:AI SRE与编码助手的市场定位差异分析 +- 问题域:AI工具的命名和营销框架如何反映对不同职业的认知偏见 +- 方法/机制:对比AI SRE和编码助手产品的命名与宣传策略 +- 结论/价值:揭示AI工具的定位揭示了构建者和买家对工作价值的认知差异 + +## Key Claims +- 编码助手被定位为增强工程师能力(赋予名字,强调合作关系) +- AI SRE被定位为替代低价值工作(通常无名字,强调消除救火) +- 这种差异反映了软件工程被视为有价值,SRE被视为阻碍性工作 +- 趋势正在从"伙伴关系"转向"泰勒制"——工人指挥AI代理群体 +- 类比既可以是杠杆,也可以是束缚 + +## Key Concepts +- [[AI SRE]]:被定位为替代SRE工作的AI工具 +- [[编码助手]]:被定位为增强工程师能力的AI工具 +- [[泰勒制]]:将工人简化为可替代单元的管理哲学 +- [[剩余原则]]:部分工作自动化后,剩余部分堆给少数人 +- [[软件工厂]]:高级控制器协调大量AI代理的架构 + +## Connections +- [[编码助手]] ← 定位 ← [[伙伴关系]] +- [[AI SRE]] ← 定位 ← [[替代工作]] +- [[Taylorism]] ← 影响 ← [[软件工厂框架]] + +## Contradictions +- 与"AI提升所有工作"观点冲突: + - 冲突点:不同工作的价值判断 + - 当前观点:AI工具定位反映职业偏见 + - 对方观点:所有工作都被平等对待 diff --git a/wiki/sources/tiktok-pm-python-django-project.md b/wiki/sources/tiktok-pm-python-django-project.md new file mode 100644 index 00000000..e397b808 --- /dev/null +++ b/wiki/sources/tiktok-pm-python-django-project.md @@ -0,0 +1,59 @@ +--- +title: "TikTok PM - Python Django 项目" +type: source +tags: [django, mariadb, mysql, python, tiktok, docker] +date: 2025-11-24 +source_file: raw/Technical/Project/TikTok PM - Python Django Project.md +--- + +## Summary +- 核心主题:TikTok电商产品管理系统的Django实现 +- 问题域:TikTok平台产品数据的抓取、存储、管理与可视化 +- 方法/机制:Django ORM + Django Admin + DRF + Docker容器化 + Bright Data异步抓取 +- 结论/价值:构建完整的电商产品数据管理后台,支持多店铺产品管理、批量导入、RESTful API + +## Key Claims +- Django Admin可通过定制化实现电商产品管理的复杂需求 +- Django REST Framework可快速构建供n8n调用的RESTful API +- Docker容器化部署可实现开发生产环境一致性 +- Bright Data异步API可实现大规模TikTok产品数据的批量抓取 + +## Key Quotes +> "Django采用"项目(Project)和应用(App)的两级结构" — Django框架的核心组织方式 + +> "Django-Q用于异步任务队列,确保Web请求不会超时" — 异步任务处理方案 + +## Key Concepts +- [[Django ORM模型设计]]:使用Python类定义数据库表结构,支持Product、ProductImage、ProductVideo、ProductVariation、ProductReview等关联模型 +- [[Django Admin定制]]:通过list_display、search_fields、list_filter、fieldsets、inlines等实现复杂管理界面 +- [[RESTful API设计]]:使用Django REST Framework的ViewSet和Router自动生成CRUD路由 +- [[Docker容器化部署]]:通过docker-compose编排Django+Gunicorn+Nginx+MySQL服务 +- [[异步任务处理]]:使用Django-Q实现Bright Data API的异步调用和轮询 +- [[富文本编辑器集成]]:集成TinyMCE实现description字段的富文本编辑 + +## Key Entities +- [[Django]]:Python Web框架,核心框架 +- [[MySQL]]:关系型数据库,存储产品数据 +- [[Docker]]:容器化部署工具 +- [[TinyMCE]]:富文本编辑器 +- [[Bright Data]]:TikTok产品数据抓取服务 +- [[Gunicorn]]:Python WSGI HTTP服务器 +- [[Nginx]]:反向代理和静态文件服务 +- [[shenwei]]:项目作者 + +## Connections +- [[Django]] ← uses ← [[Django ORM模型设计]] +- [[Django]] ← uses ← [[Django Admin定制]] +- [[Django]] ← uses ← [[RESTful API设计]] +- [[Docker]] ← orchestrates ← [[Docker容器化部署]] +- [[Django]] ← uses ← [[异步任务处理]] +- [[Django]] ← uses ← [[富文本编辑器集成]] + +## Contradictions + +## Project Structure +- tiktok_pm_project/:Django项目主配置目录 +- products/:产品应用目录,包含models.py、admin.py、views.py、serializers.py +- nginx/:Nginx配置文件目录 +- data/json/:JSON数据存储目录 +- data/html/:HTML描述文件存储目录 diff --git a/wiki/sources/todoist-task-manager.md b/wiki/sources/todoist-task-manager.md new file mode 100644 index 00000000..a53941e6 --- /dev/null +++ b/wiki/sources/todoist-task-manager.md @@ -0,0 +1,31 @@ +--- +title: "Todoist Task Manager: Agent Task Visibility" +type: source +tags: [OpenClaw, Todoist, Task Visibility, Bash Scripts] +sources: ["https://developer.todoist.com/rest/v2/"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:Agent 任务可见性 +- 问题域:如何在 Todoist 可视化长时间运行的多步骤 Agent 工作流 +- 方法/机制:任务创建/状态同步/子步骤评论/心跳检测 +- 结论/价值:复杂 Agent 任务透明可见 + +## Key Claims +- 任务状态分区(🟡进行中/🟠等待/🟢完成) +- Agent 内部推理外化到任务描述 +- 子步骤完添加为评论实时流日志 +- 心跳脚本检测停滞任务并通知用户 + +## Key Concepts +- [[Task Visibility]]:任务可见性 +- [[Agent Transparency]]:Agent 透明度 +- [[Todoist API]]:Todoist REST API + +## Connections +- [[OpenClaw]] ← 运行 ← [[Todoist Task Manager]] +- [[Todoist]] ← 平台 ← [[Todoist Task Manager]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/tong-guo-vps-nei-wang-fan-xiang-dai-li-shi-xian-yu-ming-fang-wen-nei-wang-chuan-tou.md b/wiki/sources/tong-guo-vps-nei-wang-fan-xiang-dai-li-shi-xian-yu-ming-fang-wen-nei-wang-chuan-tou.md new file mode 100644 index 00000000..f264b117 --- /dev/null +++ b/wiki/sources/tong-guo-vps-nei-wang-fan-xiang-dai-li-shi-xian-yu-ming-fang-wen-nei-wang-chuan-tou.md @@ -0,0 +1,64 @@ +--- +title: "通过VPS+内网反向代理实现域名访问内网穿透" +type: source +tags: [vps, caddy, frp, reverse-proxy, cloudflare, 内网穿透] +date: 2026-04-13 +source_file: raw/Technical/Home Office/通过VPS+内网反向代理实现域名访问内网穿透.md +--- + +## Summary +- 核心主题:通过FRP+Caddy实现从公网域名访问内网服务 +- 问题域:内网服务暴露、域名解析、反向代理配置 +- 方法/机制:frps(VPS) + frpc(内网) + Caddy反向代理 +- 结论/价值:构建完整的内网穿透访问体系 + +## Key Claims +- FRP支持TCP/HTTP/HTTPS多种映射 +- Caddy自动申请Let's Encrypt证书 +- DNS托管于Cloudflare +- 阿里云DNS解析到VPS公网IP + +## Key Concepts +- [[内网穿透]]:FRP反向隧道技术 +- [[反向代理]]:Caddy处理HTTPS +- [[Caddy]]:现代化Web服务器 +- [[FRP]]:内网穿透工具 + +## Key Entities +- [[VPS]]:公网中转服务器(192.227.222.142) +- [[Caddy]]:反向代理/HTTPS服务器 +- [[FRP]]:内网穿透工具 +- [[Cloudflare]]:DNS托管 +- [[阿里云DNS]]:域名解析 + +## Connections +- [[内网服务]] ← frpc ← [[VPS frps]] +- [[VPS]] ← Caddy反代 ← 公网 +- [[阿里云DNS]] ← 解析 ← [[Cloudflare]] + +## 架构拓扑 +``` +Internet + ↓ +VPS (192.227.222.142) + ├─ frps (端口7000) + └─ Caddy (80/443) + ↓ + 内网主机 frpc + ↓ + NAS/Ubuntu +``` + +## 域名映射表 +| 域名 | 目标端口 | 服务 | +|------|----------|------| +| nas.ishenwei.online | 15000 | NAS DSM | +| n8n.ishenwei.online | 15678 | n8n | +| grafana.ishenwei.online | 13000 | Grafana | + +## 故障排查步骤 +1. 确认frps监听端口:`ss -lntup | grep 7000` +2. 检查进程配置:`ps -ef | grep frps` +3. 验证防火墙:`sudo ufw status` +4. 检查Caddy配置:`sudo caddy validate` +5. 查看frps日志:`journalctl -u frps` \ No newline at end of file diff --git a/wiki/sources/trae-remote-dev-guide.md b/wiki/sources/trae-remote-dev-guide.md new file mode 100644 index 00000000..db454e15 --- /dev/null +++ b/wiki/sources/trae-remote-dev-guide.md @@ -0,0 +1,36 @@ +--- +title: "Trae远程开发部署指南" +type: source +tags: [Trae, Remote-SSH, Docker, Ubuntu] +date: 2026-04-14 +source_file: raw/Technical/Trae远程开发部署指南.md +--- + +## Summary +- 核心主题:通过Trae IDE远程连接Ubuntu服务器进行Docker项目开发 +- 问题域:内网开发环境的远程协作配置 +- 方法/机制:SSH免密 → Trae Remote-SSH → Attach容器/编辑宿主机 +- 结论/价值:完整的远程开发工作流,模式A(Attach容器)适合调试 + +## Key Claims +- Ubuntu 2是开发服务器(代码挂载),Ubuntu 1是生产服务器(镜像打包) +- SSH免密需要配置SSH Key并上传公钥 +- 用户必须在docker用户组中才能操作Docker +- 模式A:Attach到运行中容器(推荐调试) +- 模式B:编辑宿主机文件 + Docker CLI(适合编排) + +## Key Concepts +- [[Remote-SSH]]:通过SSH远程连接开发 +- [[Bind-Mount]]:开发环境代码修改实时生效 +- [[Docker-用户组]]:免sudo操作Docker + +## Key Entities +- [[Trae]]:支持VS Code插件生态的IDE +- [[Ubuntu]]:Linux服务器 +- [[Docker]]:容器平台 + +## Connections +- [[Git-Push-连接重置]] ← 版本控制 ← [[Trae远程开发部署]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/ubuntu-24-04-enable-ssh.md b/wiki/sources/ubuntu-24-04-enable-ssh.md new file mode 100644 index 00000000..f3ac5829 --- /dev/null +++ b/wiki/sources/ubuntu-24-04-enable-ssh.md @@ -0,0 +1,32 @@ +--- +title: "Ubuntu 24.04 enable SSH" +type: source +tags: [ssh, ubuntu] +date: 2026-04-13 +source_file: raw/Technical/Home Office/🟠Ubuntu 24.04 enable SSH.md +--- + +## Summary +- 核心主题:Ubuntu 24.04中SSH服务的启用与配置 +- 问题域:远程服务器访问管理 +- 方法/机制:安装OpenSSH Server + systemctl管理服务 + UFW防火墙配置 +- 结论/价值:Ubuntu 24.04默认使用ssh.socket socket激活机制,需了解与传统ssh.service的区别 + +## Key Claims +- Ubuntu 24.04默认使用ssh.socket激活机制,有连接请求时才启动sshd +- 安装OpenSSH Server后通过systemctl enable ssh设置开机自启 +- 必须通过UFW允许ssh流量,否则防火墙会阻止连接 +- 可通过disable ssh.socket + enable ssh.service切换回传统运行模式 + +## Key Concepts +- [[Socket激活机制]]:按需启动服务的systemd机制,ssh.socket监听22端口 +- [[SSH服务管理]]:systemd管理下的ssh.service与ssh.socket两种运行模式 + +## Key Entities +- [[Ubuntu-Server]]:SSH服务运行的操作系统 +- [[OpenSSH]]:SSH协议的开源实现 + +## Connections +- [[OpenSSH]] ← 安装于 ← [[Ubuntu-Server]] + +## Contradictions diff --git a/wiki/sources/ubuntu-an-zhuang-frp-0-65-0-x86-64-cao-zuo-bi-ji.md b/wiki/sources/ubuntu-an-zhuang-frp-0-65-0-x86-64-cao-zuo-bi-ji.md new file mode 100644 index 00000000..d5b8fd0c --- /dev/null +++ b/wiki/sources/ubuntu-an-zhuang-frp-0-65-0-x86-64-cao-zuo-bi-ji.md @@ -0,0 +1,53 @@ +--- +title: "Ubuntu 安装 FRP 0.65.0(x86_64)操作笔记" +type: source +tags: [frp, ubuntu, x86_64, 内网穿透, systemd] +date: 2026-04-13 +source_file: raw/Technical/Home Office/🟣Ubuntu 安装 FRP 0.65.0(x86_64)操作笔记.md +--- + +## Summary +- 核心主题:Ubuntu Server上安装配置FRP内网穿透客户端 +- 问题域:Linux软件安装、systemd服务管理、内网穿透配置 +- 方法/机制:下载解压 → 配置frpc.toml → systemd服务注册 +- 结论/价值:实现从公网通过VPS访问内网Ubuntu服务器 + +## Key Claims +- FRP 0.65.0支持x86_64架构 +- systemd是Linux推荐的服务管理方式 +- 软链接方案便于版本切换 + +## Key Concepts +- [[内网穿透]]:通过FRP实现外网访问内网服务 +- [[systemd]]:Linux系统服务管理器 +- [[FRP]]:高性能内网穿透工具 + +## Key Entities +- [[Ubuntu Server]]:内网服务器 +- [[FRP]]:内网穿透工具 + +## Connections +- [[Ubuntu Server]] ← 运行 ← [[FRP]] +- [[FRP]] ← 隧道 ← [[VPS]] + +## 配置参数 +| 参数 | 值 | +|------|-----| +| 安装路径 | /opt/frp/frp_0.65.0_linux_amd64 | +| 服务器地址 | 192.227.222.142 | +| 服务器端口 | 7000 | +| 服务管理 | systemd | + +## systemd服务配置 +服务文件:`/etc/systemd/system/frpc.service` +- Type=simple +- Restart=on-failure +- ExecStart=/opt/frp/frp_0.65.0_linux_amd64/frpc + +## 常用命令 +| 命令 | 说明 | +|------|------| +| sudo systemctl start frpc | 启动 | +| sudo systemctl enable frpc | 开机自启 | +| sudo systemctl status frpc | 查看状态 | +| sudo journalctl -u frpc -f | 查看日志 | \ No newline at end of file diff --git a/wiki/sources/ubuntu-fu-wu-qi-tong-guo-rsync-shi-xian-ri-chang-zeng-liang-bei-fen.md b/wiki/sources/ubuntu-fu-wu-qi-tong-guo-rsync-shi-xian-ri-chang-zeng-liang-bei-fen.md new file mode 100644 index 00000000..5d3696da --- /dev/null +++ b/wiki/sources/ubuntu-fu-wu-qi-tong-guo-rsync-shi-xian-ri-chang-zeng-liang-bei-fen.md @@ -0,0 +1,110 @@ +--- +title: "Ubuntu服务器通过rsync实现日常增量备份" +type: source +tags: [rsync, backup, ubuntu, nas, nfs, cron] +date: 2025-11-11 +source_file: raw/Technical/Home Office/🟠Ubuntu服务器通过rsync实现日常增量备份.md +--- + +## Summary +- 核心主题:使用rsync实现NAS到Ubuntu服务器的增量备份 +- 问题域:数据保护、增量同步、自动化备份 +- 方法/机制:rsync + cron定时任务 + NFS挂载 +- 结论/价值:构建工作室级数据保护体系 + +## Key Claims +- rsync只传输变化的文件,适合增量备份 +- NFS挂载需写入/etc/fstab实现永久挂载 +- _netdev参数防止开机网络未就绪时卡死 +- 备份脚本需检查挂载点防止数据写错位置 + +## Key Concepts +- [[增量备份]]:rsync只同步变化部分 +- [[NFS挂载]]:网络文件系统永久挂载 +- [[定时任务]]:cron自动化执行 +- [[数据保护]]:备份策略 + +## 备份目标 +- /var/lib/docker/volumes/(Docker卷数据) +- /etc/docker/(Docker配置) +- /home/shenwei/Docker/(自定义Docker配置) + +## 备份脚本 +```bash +#!/bin/bash +LOCKFILE="/tmp/rsync_backup.lock" +if [ -e ${LOCKFILE} ] && kill -0 `cat ${LOCKFILE}`; then + echo "备份任务已在运行中,跳过本次执行。" + exit +fi +echo $$ > ${LOCKFILE} +trap "rm -f ${LOCKFILE}" EXIT + +MOUNT_POINT="/mnt/nas_backup" +DATE=$(date +%Y-%m-%d) +DEST="$MOUNT_POINT/docker_backups/$DATE" +LOG="/var/log/rsync_backup.log" + +if ! mountpoint -q "$MOUNT_POINT"; then + echo "$(date): [错误] NAS 未挂载" >> "$LOG" + exit 1 +fi + +mkdir -p "$DEST" + +rsync -azR --delete \ + --exclude="venv/" \ + --exclude=".venv/" \ + --exclude="**/__pycache__/" \ + --exclude=".git/" \ + /var/lib/docker/volumes/ \ + /etc/docker/ \ + /home/shenwei/Docker/ \ + "$DEST/" >> "$LOG" 2>&1 +``` + +## NFS永久挂载 +在/etc/fstab添加: +``` +192.168.3.17:/volume2/backup /mnt/nas_backup nfs defaults,timeo=900,retrans=5,_netdev 0 0 +``` + +参数说明: +- timeo=900:90秒超时 +- retrans=5:重试5次 +- _netdev:等待网络就绪 + +## 定时任务 +```bash +sudo crontab -e +# 添加行: +0 3 * * * /usr/local/bin/rsync_backup.sh +``` + +每天凌晨3点执行备份。 + +## 停止rsync进程 +```bash +# 优雅停止 +sudo killall rsync + +# 强制停止 +sudo killall -9 rsync + +# 停止备份脚本 +sudo pkill -f rsync_backup.sh +``` + +## 恢复数据 +```bash +# 单文件丢失:从NAS备份目录恢复 +cp /mnt/nas_backup/.../file / + +# 系统崩溃:反向rsync +sudo rsync -aAXz /mnt/nas_backup/daily_sync/ / +``` + +## 注意事项 +- 数据库建议先mysqldump再rsync +- 返回码23/24表示部分文件权限问题,可接受 +- 先测试手动运行:`sudo nohup /usr/local/bin/rsync_backup.sh > /dev/null 2>&1 &` \ No newline at end of file diff --git a/wiki/sources/ubuntu-jin-yong-he-gai-qian-mian.md b/wiki/sources/ubuntu-jin-yong-he-gai-qian-mian.md new file mode 100644 index 00000000..a3921035 --- /dev/null +++ b/wiki/sources/ubuntu-jin-yong-he-gai-qian-mian.md @@ -0,0 +1,31 @@ +--- +title: "Ubuntu禁用合盖休眠" +type: source +tags: [ubuntu] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠Ubuntu禁用合盖休眠.md +--- + +## Summary +- 核心主题:让Ubuntu服务器在笔记本合盖后继续运行 +- 问题域:服务器笔记本合盖休眠行为控制 +- 方法/机制:修改systemd-logind配置或mask休眠target +- 结论/价值:实现合盖继续运行,适合部署为服务器的笔记本场景 + +## Key Claims +- Ubuntu 24.04中笔记本合盖休眠行为由systemd-logind控制 +- 修改HandleLidSwitch、HandleLidSwitchExternalPower、HandleLidSwitchDocked为ignore可禁用合盖休眠 +- 使用systemctl mask可以彻底禁用系统休眠功能(内核级别) + +## Key Concepts +- [[systemd-logind]]:系统登录管理器,控制电源管理行为 +- [[休眠target]]:sleep.target/suspend.target/hibernate.target/hybrid-sleep.target + +## Key Entities +- [[Ubuntu]]:Linux发行版 + +## Connections +- [[Ubuntu]] ← 配置 ← [[systemd-logind]] +- [[systemd-logind]] ← 控制 ← [[休眠target]] + +## Contradictions diff --git a/wiki/sources/ubuntu-server-ke-xue-shang-wang.md b/wiki/sources/ubuntu-server-ke-xue-shang-wang.md new file mode 100644 index 00000000..1771c818 --- /dev/null +++ b/wiki/sources/ubuntu-server-ke-xue-shang-wang.md @@ -0,0 +1,37 @@ +--- +title: "Ubuntu Server科学上网" +type: source +tags: [docker, proxychains, ubuntu, v2rayn] +date: 2026-04-13 +source_file: raw/Technical/Home Office/🟠Ubuntu Server科学上网.md +--- + +## Summary +- 核心主题:Ubuntu Server通过V2RayN代理实现终端命令、Docker、Git的网络访问 +- 问题域:Linux环境下的代理配置 +- 方法/机制:ProxyChains终端代理 + Git全局代理 + Docker Daemon代理 + 容器内代理 +- 结论/价值:针对不同场景提供差异化代理配置方案,覆盖所有网络访问路径 + +## Key Claims +- curl -x socks5h://127.0.0.1:10808可验证代理是否正常工作 +- ProxyChains通过/etc/proxychains4.conf配置,使任意命令通过代理执行 +- Git代理通过git config --global设置http.proxy和https.proxy +- Docker pull由Daemon执行,需修改systemd service的HTTP_PROXY环境变量 +- Docker容器内代理需使用宿主机网桥IP(如172.24.0.1)而非127.0.0.1 + +## Key Concepts +- [[终端代理]]:通过ProxyChains让原本不支持代理的命令走SOCKS5通道 +- [[Git代理配置]]:为Git命令设置全局代理解决clone卡顿问题 +- [[Docker代理配置]]:Docker Daemon层和容器层的代理隔离配置 + +## Key Entities +- [[V2Ray/Tuic]]:提供SOCKS5代理的客户端 +- [[Docker]]:容器平台,需要单独配置代理 +- [[Ubuntu-Server]]:运行代理配置的操作系统 + +## Connections +- [[V2Ray/Tuic]] ← 代理来源 ← [[Docker]] +- [[V2Ray/Tuic]] ← 代理来源 ← [[Git代理配置]] +- [[ProxyChains]] ← 终端代理工具 ← [[V2Ray/Tuic]] + +## Contradictions diff --git a/wiki/sources/ubuntu-yong-rustdesk-yuan-cheng-deng-lu-chu-xian-bu-neng-shi-yong-wayland-deng-lu-de-cuo-wu.md b/wiki/sources/ubuntu-yong-rustdesk-yuan-cheng-deng-lu-chu-xian-bu-neng-shi-yong-wayland-deng-lu-de-cuo-wu.md new file mode 100644 index 00000000..4a918ca4 --- /dev/null +++ b/wiki/sources/ubuntu-yong-rustdesk-yuan-cheng-deng-lu-chu-xian-bu-neng-shi-yong-wayland-deng-lu-de-cuo-wu.md @@ -0,0 +1,34 @@ +--- +title: "Ubuntu用RustDesk远程登录出现不能使用Wayland登录的错误" +type: source +tags: [ubuntu, rustdesk] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠Ubuntu用RustDesk远程登录出现不能使用Wayland登录的错误.md +--- + +## Summary +- 核心主题:RustDesk在Ubuntu 24.04上因Wayland安全限制无法远程登录 +- 问题域:远程桌面控制Ubuntu登录界面 +- 方法/机制:禁用Wayland,强制使用X11显示协议 +- 结论/价值:修改GDM3配置使RustDesk可正常接管登录界面 + +## Key Claims +- Ubuntu 24.04默认使用Wayland显示协议,出于安全设计严格限制外部程序在用户未登录状态下获取屏幕控制权 +- 禁用Wayland后,Ubuntu登录界面强制使用X11加载,RustDesk后台服务能够识别X11窗口并与其交互 +- X11的稳定性与权限开放度目前仍优于Wayland,适合需要频繁远程桌面运维的场景 + +## Key Concepts +- [[Wayland]]:Ubuntu 24.04默认的显示协议,出于安全考虑限制远程控制 +- [[X11]]:传统显示协议,兼容性更好,允许RustDesk等远程控制软件工作 +- [[GDM3]]:GNOME显示管理器,控制登录界面 + +## Key Entities +- [[RustDesk]]:开源远程桌面软件 +- [[Ubuntu]]:Linux发行版 + +## Connections +- [[Ubuntu]] ← 使用 ← [[GDM3]] +- [[GDM3]] ← 配置禁用 ← [[Wayland]] +- [[RustDesk]] ← 依赖 ← [[X11]] + +## Contradictions diff --git a/wiki/sources/useful-prompt-lib.md b/wiki/sources/useful-prompt-lib.md new file mode 100644 index 00000000..4e569c0b --- /dev/null +++ b/wiki/sources/useful-prompt-lib.md @@ -0,0 +1,30 @@ +--- +title: "Useful Prompt Lib" +type: source +tags: [Claude, AI, Prompt] +date: 2026-04-14 +source_file: raw/Technical/Useful Prompt Lib.md +--- + +## Summary +- 核心主题:Claude AI Prompt库精选推荐 +- 问题域:TikTok跨境电商业务需要哪些Prompt +- 结论/价值:Babel's broadcasts多语言本地化、Review classifier评论分类、Data organizer数据整理 + +## Key Claims +- Babel's broadcasts:适合TikTok视频脚本多语言本地化 +- Review classifier:TikTok店铺/广告评论自动化分类 +- Data organizer:非结构化产品信息转JSON,对接自动化工作流 + +## Key Concepts +- [[Prompt库]]:AI提示词模板集合 +- [[多语言本地化]]:内容适配不同语言市场 + +## Key Entities +- [[Claude]]:AI模型 + +## Connections +- [[bu-tan-ji-shu]] ← AI应用 ← [[Claude-Prompt库]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/vibe-coding-cn.md b/wiki/sources/vibe-coding-cn.md new file mode 100644 index 00000000..5554d17d --- /dev/null +++ b/wiki/sources/vibe-coding-cn.md @@ -0,0 +1,59 @@ +--- +title: "GitHub 上 5000 人收藏的 Vibe Coding 神级指南" +type: source +tags: [vibe-coding, ai-programming, github] +sources: ["https://mp.weixin.qq.com/s/QMPMSGW6XXk8L-yx4ujQcw"] +last_updated: 2025-12-30 +--- + +## Summary + +- 核心主题:Vibe Coding(氛围编程)方法论与vibe-coding-cn中文资源库 +- 问题域:如何利用AI编程工具高效完成软件开发 +- 方法/机制:规划驱动+上下文固定+AI结对执行,将想法转化为可维护代码 +- 结论/价值:为中文开发者提供Vibe Coding体系化工具链、提示词库和开发经验 + +## Key Claims + +- Vibe Coding = 规划驱动 + 上下文固定 + AI结对执行,让从想法到可维护代码变成可审计流水线 +- AI编程工具(Cursor、Windsurf、Trae)承担体力活,开发者扮演导演角色 +- Karpathy几乎不写代码,只负责调整氛围,代码自动生成 +- 规划就是一切,防止AI因理解偏差导致项目逻辑混乱 + +## Key Quotes + +> "Vibe Coding说白了就是开发个应用不再像程序员一样,苦哈哈地写每一行代码,而是化身为导演。" — 文章定义 + +> "只需要保持一种感觉,这种感觉可能是对产品逻辑、用户流程、审美和交互的把握,剩下的体力活全交给AI编程工具。" — 核心理念 + +## Key Concepts + +- [[Vibe Coding]](氛围编程):AI辅助编程范式,开发者聚焦产品和设计,AI负责代码实现 +- [[规划驱动]]:AI写代码前必须有清晰技术选型、实施规划和模块化设计 +- [[上下文固定]]:保持AI对项目上下文的理解一致性 +- [[AI结对执行]]:人类与AI协作完成开发任务 +- [[提示词优化]]:自然语言清晰定义需求,保持上下文一致 + +## Key Entities + +- [[Cursor]]:AI编程IDE +- [[Windsurf]]:AI编程IDE +- [[Trae]]:AI编程IDE +- [[Karpathy]]:AI领域专家,Vibe Coding概念提出者 +- [[vibe-coding-cn]]:中文Vibe Coding开源项目,GitHub开源资源库 + +## Connections + +- [[Vibe Coding]] ← 工具 ← [[Cursor]] +- [[Vibe Coding]] ← 工具 ← [[Windsurf]] +- [[Vibe Coding]] ← 工具 ← [[Trae]] +- [[vibe-coding-cn]] ← 包含 ← [[Vibe Coding方法论]] +- [[vibe-coding-cn]] ← 包含 ← [[AI编程资源]] +- [[vibe-coding-cn]] ← 包含 ← [[提示词优化技巧]] + +## Contradictions + +- 与传统编程方法冲突: + - 冲突点:代码所有权和责任 + - 当前观点:开发者扮演导演,AI生成代码 + - 对方观点:开发者应完全掌控每一行代码 diff --git a/wiki/sources/vibe-coding-experience-collection.md b/wiki/sources/vibe-coding-experience-collection.md new file mode 100644 index 00000000..83464d7d --- /dev/null +++ b/wiki/sources/vibe-coding-experience-collection.md @@ -0,0 +1,38 @@ +--- +id: vibe-coding-experience-collection +title: "Vibe Coding 经验收集" +type: source +tags: [vibe-coding, ai-agent, prompt-engineering] +sources: [] +last_updated: 2026-04-14 +--- + +## Summary +- 核心主题:Vibe Coding 实践经验和最佳实践收集 +- 问题域:如何有效使用 AI 编码 Agent 提升开发效率 +- 方法/机制:伪代码前置、点线面迭代、文件头注释文档化 +- 结论/价值:AI 辅助编程的核心是"验证代码按正确逻辑运行"而非"看懂代码" + +## Key Claims +- 需求 → 伪代码 → 代码的流程可显著提升 AI 生成质量 +- 点 → 线 → 体的逐级迭代是批量任务执行的有效方法 +- 文件头注释降低认知负载,参考 Claude Skill 格式 +- 未来软件工程核心是"验证代码"而非"看懂代码" + +## Key Quotes +> "我是把设计文档写得很细,包括service层的具体逻辑都用伪代码写了,然后交给AI,一遍直出,再用另一个AI review一遍" — Vibe Coding 实践经验 + +> "随着软件规模不断扩大,靠人眼去'看代码'不仅无法应对增长的复杂度...未来的软件工程核心不是'看懂代码',而是'验证代码按正确逻辑运行'" — @dogejustdoit + +## Key Concepts +- [[Vibe-Coding]]:一种 AI 辅助编程范式,通过自然语言描述需求,AI Agent 自动生成代码 +- [[Agent模式]]:AI Agent 自动执行命令的工作模式 + +## Key Entities +- [[Claude-Code-Templates]]:Claude Code 模板集合,提供 Skill、Agent、MCP 等模板 + +## Connections +- [[Vibe-Coding]] ← 依赖 ← [[Agent模式]] +- [[Claude-Code-Templates]] → 提供 → [[Vibe-Coding]] + +## Contradictions diff --git a/wiki/sources/vibe-kanban-opencode-ubuntu-guide.md b/wiki/sources/vibe-kanban-opencode-ubuntu-guide.md new file mode 100644 index 00000000..23aaff04 --- /dev/null +++ b/wiki/sources/vibe-kanban-opencode-ubuntu-guide.md @@ -0,0 +1,38 @@ +--- +id: vibe-kanban-opencode-ubuntu-guide +title: "Vibe-Kanban + OpenCode 在 Ubuntu Server 上安装与管理指南" +type: source +tags: [npm, npx, pm2, ubuntu, vibe-coding, vibe-kanban] +sources: [] +last_updated: 2026-04-14 +--- + +## Summary +- 核心主题:Ubuntu Server 环境下 Vibe-Kanban 与 OpenCode 的完整安装与进程管理 +- 问题域:AI 编码 Agent 的本地化部署与运维管理 +- 方法/机制:通过 nvm 安装 Node 20,使用 pm2 管理进程,确保非 root 用户权限正确 +- 结论/价值:提供可复用的生产环境部署指南 + +## Key Claims +- Vibe-Kanban 和 OpenCode 必须以非 root 用户运行,避免权限问题 +- Node 版本必须使用 20,兼容最新工具链 +- pm2 可实现进程自动重启和开机自启 +- `/var/tmp/vibe-kanban` 和 `~/.vibe-kanban` 目录权限必须属于运行用户 + +## Key Concepts +- [[Vibe-Kanban]]:AI 编码 Agent 任务看板,支持多 Agent 协作 +- [[OpenCode]]:开源 AI 编码 Agent,提供终端界面和桌面应用 +- [[PM2]]:Node.js 进程管理器,支持进程守护和负载均衡 +- [[nvm]]:Node.js 版本管理器,支持多版本共存 +- [[Node-js]]:JavaScript 运行时环境 + +## Key Entities +- [[BloopAI]]:Vibe-Kanban 背后的公司,提供开源 AI 编码工具 + +## Connections +- [[Vibe-Kanban]] ← 安装依赖 ← [[Node-js]] +- [[OpenCode]] ← 安装依赖 ← [[Node-js]] +- [[Vibe-Kanban]] ← 使用 ← [[PM2]] +- [[Node-js]] ← 安装工具 ← [[nvm]] + +## Contradictions diff --git a/wiki/sources/vibe-kanban-ubuntu-install.md b/wiki/sources/vibe-kanban-ubuntu-install.md new file mode 100644 index 00000000..b861fd14 --- /dev/null +++ b/wiki/sources/vibe-kanban-ubuntu-install.md @@ -0,0 +1,34 @@ +--- +id: vibe-kanban-ubuntu-install +title: "在 Ubuntu 上安装 Vibe-Kanban" +type: source +tags: [npm, npx, pm2, ubuntu, vibe-coding, vibe-kanban] +sources: [] +last_updated: 2026-04-14 +--- + +## Summary +- 核心主题:在 Ubuntu 系统上安装和配置 Vibe-Kanban +- 问题域:Node.js 环境搭建和 AI 编码 Agent 的进程管理 +- 方法/机制:使用 npx 直接运行,通过 PM2 实现后台进程管理和开机自启 +- 结论/价值:Vibe-Kanban 可绑定到指定端口(PORT 环境变量)并支持远程访问 + +## Key Claims +- Vibe-Kanban 默认绑定随机端口,可通过 `PORT` 环境变量指定固定端口 +- 使用 `HOST=0.0.0.0` 可允许远程访问 +- PM2 可实现进程守护、自动重启和开机自启 +- Vibe Kanban 默认使用 `--dangerously-skip-permissions` 标志运行,任务在独立 git worktree 中隔离执行 + +## Key Concepts +- [[Vibe-Kanban]]:AI 编码 Agent 任务看板工具 +- [[PM2]]:Node.js 进程管理器,用于守护 vibe-kanban 进程 +- [[Node-js]]:JavaScript 运行时环境 + +## Key Entities +- [[BloopAI]]:Vibe-Kanban 的开发公司 + +## Connections +- [[Vibe-Kanban]] ← 安装方式 ← [[Node-js]] +- [[Vibe-Kanban]] ← 进程管理 ← [[PM2]] + +## Contradictions diff --git a/wiki/sources/wang-jian-rax50-lu-you-qi-shua-mei-lin-gu-jian-yu-ke-xue-shang-wang-cha-jian-an-zhuang-jiao-cheng.md b/wiki/sources/wang-jian-rax50-lu-you-qi-shua-mei-lin-gu-jian-yu-ke-xue-shang-wang-cha-jian-an-zhuang-jiao-cheng.md new file mode 100644 index 00000000..691effeb --- /dev/null +++ b/wiki/sources/wang-jian-rax50-lu-you-qi-shua-mei-lin-gu-jian-yu-ke-xue-shang-wang-cha-jian-an-zhuang-jiao-cheng.md @@ -0,0 +1,34 @@ +--- +title: "网件RAX50路由器刷梅林固件与科学上网插件安装教程" +type: source +tags: [RAX50, Merlin, Clash, 路由器] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠网件RAX50路由器刷梅林固件与科学上网插件安装教程.md +--- + +## Summary +- 核心主题:网件RAX50刷梅林固件并配置科学上网 +- 问题域:路由器固件刷入、插件安装、节点配置 +- 结论/价值:完整刷机流程+JFFS双清+MerlinClash策略组配置 + +## Key Claims +- 必须先刷.chk过渡固件,再刷.w正式梅林固件 +- MerlinClash支持自动延迟测试和策略组分流 +- 两个科学上网插件不能同时开启 +- JFFS双清清理缓存确保固件干净 + +## Key Concepts +- [[Merlin固件]]:华硕路由器第三方固件改良版 +- [[MerlinClash]]:基于Clash的高级分流插件 +- [[JFFS双清]]:清理文件系统缓存 +- [[故障转移]]:连接故障时自动切换备用节点 + +## Key Entities +- [[RAX50]]:网件路由器 +- [[Merlin-Clash]]:科学上网插件 + +## Connections +- [[群晖NAS科学上网]] ← 替代方案 ← [[RAX50梅林固件]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/wo-yong-gemini-3-yi-qi-zuo-liao-10-ge-ying-yong-fu-jiao-cheng.md b/wiki/sources/wo-yong-gemini-3-yi-qi-zuo-liao-10-ge-ying-yong-fu-jiao-cheng.md new file mode 100644 index 00000000..d3756872 --- /dev/null +++ b/wiki/sources/wo-yong-gemini-3-yi-qi-zuo-liao-10-ge-ying-yong-fu-jiao-cheng.md @@ -0,0 +1,28 @@ +--- +title: "我用 Gemini 3 一口气做了 10 个应用,附教程" +type: source +tags: [gemini, gemini-3, application, prompt-engineering] +date: 2025-12-18 +source_file: raw/AI/我用 Gemini 3 一口气做了 10 个应用,附教程.md +--- + +## Summary +- 核心主题:Gemini 3多应用开发方法论 +- 问题域:AI应用开发、提示词工程 +- 方法/机制:约束模型思考 + 设计输出容器 +- 结论/价值:三步法快速构建AI应用 + +## Key Claims +- 垂直场景输入可限制AI词汇范围,提高输出质量 +- 提示词和MCP可将输入词汇扩展为结构化内容 +- 前端代码可用于可视化模型输出内容 + +## Key Concepts +- [[Gemini-3应用开发方法论]]:三步法构建AI应用 +- [[结构化输出]]:约束大模型按特定格式输出 + +## Key Entities +- [[Gemini-3]]:Google的多模态AI模型 + +## Connections +- [[Gemini-3]] ← 应用开发 ← [[结构化输出]] \ No newline at end of file diff --git a/wiki/sources/x-account-analysis.md b/wiki/sources/x-account-analysis.md new file mode 100644 index 00000000..81c2160c --- /dev/null +++ b/wiki/sources/x-account-analysis.md @@ -0,0 +1,30 @@ +--- +title: "X Account Analysis" +type: source +tags: [OpenClaw, X, Twitter, Analytics, Bird Skill] +date: 2026-04-14 +--- + +## Summary +- 核心主题:X 账户分析 +- 问题域:如何获取账户的定性分析(而非只是统计) +- 方法/机制:Bird Skill 抓取推文,AI 分析质量模式 +- 结论/价值:免费替代 $10-$50 的订阅分析服务 + +## Key Claims +- 统计 vs 定性分析:找出病毒式传播模式 +- 分析:什么话题带来最多参与度 +- 诊断:为什么有些帖子 1000+ 赞而有些 <5 赞 +- 无需付费订阅网站 + +## Key Concepts +- [[X Account Analysis]]:X 账户分析 +- [[Qualitative Analysis]]:定性分析 +- [[Bird Skill]]:X 数据抓取工具 + +## Connections +- [[OpenClaw]] ← 运行 ← [[X Account Analysis]] +- [[Bird Skill]] ← 工具 ← [[X Account Analysis]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/x-twitter-automation.md b/wiki/sources/x-twitter-automation.md new file mode 100644 index 00000000..d810095b --- /dev/null +++ b/wiki/sources/x-twitter-automation.md @@ -0,0 +1,31 @@ +--- +title: "X/Twitter Automation from Chat" +type: source +tags: [OpenClaw, X, Twitter, TweetClaw, Automation] +sources: ["https://github.com/Xquik-dev/tweetclaw"] +date: 2026-04-14 +--- + +## Summary +- 核心主题:X/Twitter 自然语言自动化 +- 问题域:如何在单一对话界面完成所有 X 操作 +- 方法/机制:TweetClaw 插件连接 X API,发推/回复/喜欢/转发/DM/搜索/数据提取/抽奖 +- 结论/价值:通过聊天完全控制 X/Twitter + +## Key Claims +- 发布和互动:发推/回复/喜欢/转发/关注/取消关注/DM +- 搜索和提取:搜索推文/用户/提取粉丝点赞用户 +- 抽奖:从推文参与中随机选择中奖者 +- 监控:监视账号新推文/粉丝变化通知 + +## Key Concepts +- [[TweetClaw]]:X/Twitter 自动化插件 +- [[X Automation]]:X 自动化 +- [[Giveaway Automation]]:抽奖自动化 + +## Connections +- [[OpenClaw]] ← 运行 ← [[X Twitter Automation]] +- [[TweetClaw]] ← 工具 ← [[X Twitter Automation]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/xi-tong-ti-shi-ci-gou-jian-yuan-ze.md b/wiki/sources/xi-tong-ti-shi-ci-gou-jian-yuan-ze.md new file mode 100644 index 00000000..6d3dd07b --- /dev/null +++ b/wiki/sources/xi-tong-ti-shi-ci-gou-jian-yuan-ze.md @@ -0,0 +1,26 @@ +--- +title: "系统提示词构建原则" +type: source +tags: [prompt-engineering, system-prompt, ai-development] +date: 2025-12-30 +source_file: raw/AI/系统提示词构建原则.md +--- + +## Summary +- 核心主题:AI助手的系统提示词构建原则 +- 问题域:AI开发规范、提示词工程 +- 方法/机制:身份准则、沟通规范、技术编码、安全防护 +- 结论/价值:系统化的AI开发方法论 + +## Key Claims +- 严格遵守项目现有约定,优先分析周围代码和配置 +- 绝不假设库或框架可用,务必先验证 +- 保持一致性,不轻易改变已设定的行为模式 +- 所有更改必须是完整的代码行 + +## Key Concepts +- [[系统提示词构建]]:AI助手行为准则定义 +- [[结构化输出]]:大模型按特定格式输出信息 + +## Connections +- [[系统提示词构建]] ← 依赖 ← [[变量命名规范]] \ No newline at end of file diff --git a/wiki/sources/xiang-xi-ling-shi-bu-shu-da-mo-xing.md b/wiki/sources/xiang-xi-ling-shi-bu-shu-da-mo-xing.md new file mode 100644 index 00000000..e695db9b --- /dev/null +++ b/wiki/sources/xiang-xi-ling-shi-bu-shu-da-mo-xing.md @@ -0,0 +1,30 @@ +--- +title: "详细!离线部署大模型:ollama+deepseek+open-webui安装使用方法及常见问题解决" +type: source +tags: [ollama, deepseek, open-webui, local-llm, deployment] +date: 2025-03-14 +source_file: raw/AI/详细!离线部署大模型:ollama+deepseek+open-webui安装使用方法及常见问题解决 1.md +--- + +## Summary +- 核心主题:ollama + DeepSeek + open-webui 本地大模型部署 +- 问题域:本地LLM部署、离线运行 +- 方法/机制:Docker容器化、API调用、可视化界面 +- 结论/价值:完整的本地大模型部署解决方案 + +## Key Claims +- ollama提供简单便捷的本地LLM运行框架 +- 4GB RAM运行1.5B模型,16GB RAM运行7B模型 +- open-webui提供图形化界面支持聊天机器人和本地知识库 +- 支持从HF和魔塔社区下载第三方模型 + +## Key Entities +- [[Ollama]]:开源本地大语言模型运行框架 +- [[DeepSeek]]:深度求索大模型 +- [[open-webui]]:开源AI Web界面 +- [[Docker]]:容器化部署工具 + +## Connections +- [[Ollama]] ← 模型运行 ← [[DeepSeek]] +- [[open-webui]] ← API集成 ← [[Ollama]] +- [[Docker]] ← 容器化 ← [[Ollama]] \ No newline at end of file diff --git a/wiki/sources/yang-long-xia-5-tian.md b/wiki/sources/yang-long-xia-5-tian.md new file mode 100644 index 00000000..be495033 --- /dev/null +++ b/wiki/sources/yang-long-xia-5-tian.md @@ -0,0 +1,50 @@ +--- +title: "养龙虾5天血泪史:我的AI Agent为什么总失忆?OpenClaw 记忆调试全记录" +type: source +tags: [AI, Agent, OpenClaw, 记忆管理, Context-Window] +date: 2026-04-02 +source_file: raw/微信公众号/养龙虾5天血泪史:我的AI Agent为什么总失忆?OpenClaw 记忆调试全记录.md +--- + +## Summary +- 核心主题:OpenClaw Agent记忆问题的系统化调试经历 +- 问题域:长对话后上下文丢失、搜索返回垃圾结果、Agent找到信息但不使用、模型切换失忆 +- 方法/机制:5天逐日排查,最终建立完整的记忆管理规范 +- 结论/价值:记忆系统需要写入纪律、检索测试、交接协议;压缩不是敌人,未写入的上下文才是 + +## Key Claims +- 自动加载的文件只有7个:AGENTS.md/SOUL.md/TOOLS.md/IDENTITY.md/USER.md/HEARTBEAT.md/MEMORY.md +- 启动序列必须放在AGENTS.md顶部,否则Agent不会读取 +- 写入纪律比读取纪律更重要:临时上下文必须写盘才能在压缩中存活 +- 混合搜索(BM25+向量+重排序)优于纯语义搜索 +- 交接协议解决模型切换时上下文丢失问题 + +## Key Quotes +> "真正的修复不是添加更多文件。而是移除那些什么都不做的文件。" — 比利哥 + +> "压缩不是敌人。未写入的上下文才是。" — 比利哥 + +> "你花一个小时配置每日定时任务,切换模型后,Agent表现得像我们从未交谈过。" — 比利哥 + +## Key Concepts +- [[Context-Compaction]]:对话压缩机制,摘要抓住要点但丢失细节 +- [[Memory-Flush]]:压缩前将重要上下文写入磁盘 +- [[混合搜索]]:BM25(关键词)+ 向量(语义)+ 重排序 +- [[交接协议]]:模型切换前将当前上下文写入每日日志 +- [[写入纪律]]:任务结果必须记录到磁盘 +- [[检索测试]]:植入标记并跨会话验证Agent能找到 + +## Key Entities +- [[OpenClaw]]:AI Agent操作系统 +- [[星辉]](xinghui):OpenClaw Agent,本文主要调试对象 +- [[QMD]]:本地Markdown搜索引擎,支持混合搜索 +- [[MiniMax-M2.7]]:200K context主力模型 +- [[比利哥]]:作者 + +## Connections +- [[yang-xia-ri-ji-2]] ← 深化 ← [[yang-long-xia-5-tian]](本文是yang-xia-ri-ji-2记忆思想的详细实践版) +- [[yang-xia-ri-ji-3]] ← 技术支撑 ← [[yang-long-xia-5-tian]](Obsidian是记忆的物理载体) +- [[yang-xia-ri-ji-4]] ← 问题关联 ← [[yang-long-xia-5-tian]](Context Limit与记忆管理密切相关) + +## Contradictions +- 与"Agent会自动记住重要信息"假设冲突:需要主动设计写入纪律和检索触发 diff --git a/wiki/sources/yang-xia-ri-ji-1.md b/wiki/sources/yang-xia-ri-ji-1.md new file mode 100644 index 00000000..312492df --- /dev/null +++ b/wiki/sources/yang-xia-ri-ji-1.md @@ -0,0 +1,37 @@ +--- +title: "养虾日记1:我用 OpenClaw 管了 28 万张照片" +type: source +tags: [AI, Agent, OpenClaw, 照片管理] +date: 2026-03-31 +source_file: raw/微信公众号/养虾日记1:我用 OpenClaw 管了 28 万张照片:一次真实的多设备照片整理实战.md +--- + +## Summary +- 核心主题:用AI Agent管理海量照片的实战经验 +- 问题域:多设备照片备份混乱、重复文件多、20年积累难整理 +- 方法/机制:OpenClaw分析需求 → 制定方案 → 定时批次执行 +- 结论/价值:AI Agent的价值是思维升级而非单点能力提升 + +## Key Claims +- AI Agent通过提问帮助用户把模糊需求变成可执行方案 +- 28万张照片分8批次每天凌晨自动处理,无需人工介入 +- AI Agent的核心价值是"把模糊想法变成清晰结构,把大任务拆成可执行批次" + +## Key Quotes +> "AI Agent对我来说真正的价值:不是某个单点能力的提升,而是思维方式的升级。" — 比利哥 + +## Key Concepts +- [[需求澄清]]:AI通过提问把模糊需求变成精确方案 +- [[批次处理]]:大任务拆分为可管理的定时批次 +- [[安全删除策略]]:待删文件移至To-Be-Deleted目录而非直接删除 + +## Key Entities +- [[OpenClaw]]:AI Agent操作系统,本文的核心工具 +- [[比利哥]]:作者,养虾(研究AI Agent应用)的实践者 + +## Connections +- [[养虾日记2]] ← 技术演进 ← [[yang-xia-ri-ji-1]](同一作者,Agent记忆问题进阶) +- [[bu-tan-ji-shu]] ← 实证 ← [[yang-xia-ri-ji-1]](案例印证品味+执行力) + +## Contradictions +- 与"给AI简单指令"思维冲突:本文强调AI需要通过对话澄清需求,而非直接下达去重命令 diff --git a/wiki/sources/yang-xia-ri-ji-2.md b/wiki/sources/yang-xia-ri-ji-2.md new file mode 100644 index 00000000..d438be3d --- /dev/null +++ b/wiki/sources/yang-xia-ri-ji-2.md @@ -0,0 +1,43 @@ +--- +title: "养虾日记2:让Agent更懂你:OpenClaw + Self-Improving 复盘" +type: source +tags: [AI, Agent, OpenClaw, Self-Improving, 记忆系统] +date: 2026-04-01 +source_file: raw/微信公众号/养虾日记2:让Agent更懂你:OpenClaw + Self-Improving 复盘实战案例分享.md +--- + +## Summary +- 核心主题:解决AI Agent每次对话都是白纸的记忆问题 +- 问题域:Agent没有长期记忆,每次对话需要重新描述上下文 +- 方法/机制:三层记忆架构(短期记忆文件 + 长期向量数据库 + self-improving复盘) +- 结论/价值:错误只犯一次,第二次就知道怎么做——self-improving让Agent持续进化 + +## Key Claims +- 双层记忆架构:memory文件管上下文,LanceDB管长期知识 +- self-improving skill让每个错误只犯一次,Recurrence-Count是核心指标 +- 每日23:00定时复盘推动Agent持续改进 + +## Key Quotes +> "三层各司其职:每日文件管上下文,向量数据库管知识,self-improving 管成长。" — 比利哥 + +> "错误只犯一次,第二次就知道怎么做对。" — 比利哥 + +## Key Concepts +- [[三层记忆架构]]:短期文件(每日对话)+ 长期向量(LanceDB)+ self-improving复盘 +- [[Self-Improving Skill]]:结构化经验记录系统,固定格式让Agent能检索学习 +- [[Pattern-Key]]:错误分类标签,重复出现是系统性问题的信号 +- [[Recurrence-Count]]:重复次数指标,>1说明需要系统性解决 + +## Key Entities +- [[OpenClaw]]:AI Agent操作系统 +- [[星辉]](xinghui):OpenClaw Agent之一 +- [[LanceDB]]:向量数据库,用于长期记忆存储 +- [[比利哥]]:作者 + +## Connections +- [[yang-xia-ri-ji-1]] → 技术基础 → [[yang-xia-ri-ji-2]](同一作者,从照片管理到记忆系统) +- [[养虾日记4]] ← 问题延续 ← [[yang-xia-ri-ji-2]](Context Limit问题可能与记忆系统相关) +- [[养龙虾5天血泪史]] ← 深化 ← [[yang-xia-ri-ji-2]](更全面的记忆调试经验) + +## Contradictions +- 与"Agent应该天生记住一切"假设冲突:实际需要主动设计记忆系统 diff --git a/wiki/sources/yang-xia-ri-ji-3.md b/wiki/sources/yang-xia-ri-ji-3.md new file mode 100644 index 00000000..8ecf1324 --- /dev/null +++ b/wiki/sources/yang-xia-ri-ji-3.md @@ -0,0 +1,46 @@ +--- +title: "养虾日记3:用 Obsidian + Gitea 为 AI 助手构建持久化笔记系统" +type: source +tags: [AI, Agent, Obsidian, Gitea, 知识管理] +date: 2026-04-09 +source_file: raw/微信公众号/养虾日记3:用 Obsidian + Gitea 为 AI 助手构建持久化笔记系统.md +--- + +## Summary +- 核心主题:用Obsidian做知识库、Gitea做版本控制、OpenClaw做写入接口 +- 问题域:AI助手输出只在对话中存在,对话结束内容消失 +- 方法/机制:AI执行任务时顺手写笔记到Obsidian,Git版本控制历史 +- 结论/价值:把AI变成会自动整理笔记的实习生,做完事顺手记录更新 + +## Key Claims +- OpenClaw的obsidian skill支持write/append/read/search/update五种操作 +- Agent Archive(按Agent分目录)vs Knowledge Base(公共知识)的分层结构 +- Git历史实现"任何时候都能回溯",commit message记录变更来源 +- Karpathy的LLM Wiki思路:增量构建持久化Wiki,页面互相链接,知识越积越厚 + +## Key Quotes +> "本质上是把AI变成了一个'会自动整理笔记的实习生'——它做完事,就会顺手把记录更新好。" — 比利哥 + +> "RAG模式是'每次从零检索',知识不积累;而LLM Wiki是让AI增量构建和维护一个持久化的Wiki。" — Karpathy + +## Key Concepts +- [[LLM Wiki]]:AI增量构建和维护持久化Wiki,区别于RAG每次从零检索 +- [[Obsidian Skill]]:OpenClaw的笔记操作技能(write/append/read/search/update) +- [[Agent Archive]]:单一Agent的私有笔记目录 +- [[Knowledge Base]]:跨Agent共用的整理后知识 +- [[Graph View]]:Obsidian的图谱视图,用于发现孤岛页面和知识盲区 + +## Key Entities +- [[OpenClaw]]:AI Agent操作系统 +- [[Obsidian]]:笔记管理工具 +- [[Gitea]]:自建Git服务,用于版本控制 +- [[Karpathy]]:AI研究员,提出LLM Wiki概念 +- [[星枢]]/[[星辉]]/[[星曜]]:OpenClaw Agents,各有专属Archive + +## Connections +- [[yang-xia-ri-ji-2]] ← 记忆层 ← [[yang-xia-ri-ji-3]](Obsidian笔记是记忆的物理载体) +- [[bu-tan-ji-shu]] ← 工具支撑 ← [[yang-xia-ri-ji-3]](端到端能力需要持久化笔记系统) +- [[养龙虾5天血泪史]] ← 技术细节 ← [[yang-xia-ri-ji-3]](两文都强调版本控制和写回纪律) + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/yang-xia-ri-ji-4.md b/wiki/sources/yang-xia-ri-ji-4.md new file mode 100644 index 00000000..02277670 --- /dev/null +++ b/wiki/sources/yang-xia-ri-ji-4.md @@ -0,0 +1,42 @@ +--- +title: "养虾日记4:一次「Context Limit Exceeded」错误排查" +type: source +tags: [AI, Agent, OpenClaw, Debug, Context-Window] +date: 2026-04-10 +source_file: raw/微信公众号/养虾日记4: 一次「Context Limit Exceeded」错误排查:我以为是小问题,结果踩了大坑.md +--- + +## Summary +- 核心主题:OpenClaw Agent Context Limit问题的排查过程 +- 问题域:星枢Telegram频道报Context Limit Exceeded,修改compaction配置无效 +- 方法/机制:日志分析 → 发现模型被切换到deepseek-reasoner(16K context)→ 还原正确配置 +- 结论/价值:错误信息可能是表面意思,真正问题藏在配置层级深处 + +## Key Claims +- 星枢的Telegram channel绑定到了只有16K context的deepseek-reasoner模型 +- OpenClaw配置分层:Global Config(openclaw.json)vs Agent/Channel Specific Config +- 解决后建议:让星枢Telegram用回MiniMax-M2.7(200K context) + +## Key Quotes +> "不要默认认为错误信息就是表面意思。" — 比利哥 + +> "日志真的有用:Gateway日志把问题写得明明白白,只是我自己没仔细看。" — 比利哥 + +## Key Concepts +- [[Context-Window]]:模型的上下文窗口大小,影响对话长度 +- [[Compaction]]:对话压缩机制,预留一半token给compaction(safeguard模式) +- [[Fallback机制]]:模型回退触发条件(503/429/Timeout/配置错误) +- [[配置分层]]:Global Config vs Agent/Channel Specific Config + +## Key Entities +- [[OpenClaw]]:AI Agent操作系统 +- [[星枢]](xingshu):OpenClaw主Agent +- [[deepseek-reasoner]]:16K context的推理模型(问题根源) +- [[MiniMax-M2.7]]:200K context的主力模型 + +## Connections +- [[yang-xia-ri-ji-2]] ← 问题关联 ← [[yang-xia-ri-ji-4]](Context Limit与记忆系统紧密相关) +- [[养龙虾5天血泪史]] ← 深化 ← [[yang-xia-ri-ji-4]](更全面的内存问题调试经验) + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/yang-xia-ri-ji-5.md b/wiki/sources/yang-xia-ri-ji-5.md new file mode 100644 index 00000000..2cd5b622 --- /dev/null +++ b/wiki/sources/yang-xia-ri-ji-5.md @@ -0,0 +1,43 @@ +--- +title: "养虾日记5:深夜与苏轼聊AI" +type: source +tags: [AI, Agent, Skill, 苏东坡, 思维方式] +date: 2026-04-11 +source_file: raw/微信公众号/养虾日记5:深夜与苏轼聊AI,他说:被浪打下去还能爬起来的才叫风流.md +--- + +## Summary +- 核心主题:用"女娲·Skill"蒸馏苏东坡的核心思维框架,构建可对话的数字导师 +- 问题域:如何用AI复活历史人物,让他在日常生活中提供思维咨询 +- 方法/机制:6个并行Agent从6个维度采集信息 → 提炼心智模型 → 构建.skill文件 +- 结论/价值:每个人的Skill都是一个认知操作系统,用别人的脑子思考自己的人生 + +## Key Claims +- "造人"是蒸馏过程:从大量公开信息中提取3-7个核心心智模型、5-10条决策启发式 +- 苏东坡6个核心心智模型:进退由时/此心安处/辞达而已/以时受力/自出新意/天人合一 +- 女娲工作流:Phase 0.5→1(6并行Agent)→1.5(Review)→2(提炼)→2.5(确认)→3(构建)→4(验证)→5(双Agent精炼) + +## Key Quotes +> "真正风流的人,不是站在浪尖上的人,而是被浪打下去、还能爬起来的人。" — 苏东坡(蒸馏人格) + +> "用AI放大人类历史上那些最强大的脑子,让它们成为你日常的思维顾问。" — 女娲.skill + +## Key Concepts +- [[Skill蒸馏]]:从大量信息中提取核心心智模型,构建可运行的AI Skill +- [[认知操作系统]]:每个Skill是一个思维框架,用于分析具体问题 +- [[女娲·Skill]]:造人术框架,用于创建人格化AI Skill +- [[苏东坡心智模型]]:进退由时/此心安处/辞达而已/以时受力/自出新意/天人合一 + +## Key Entities +- [[苏东坡]](Su Dongpo):1037-1101,北宋文学家,被蒸馏成AI Skill +- [[女娲·Skill]]:Skill创建框架 +- [[乔布斯.skill]]:另一个已创建的思维顾问Skill +- [[比利哥]]:作者 +- [[芒格]]/[[费曼]]/[[塔勒布]]:计划蒸馏的目标人物 + +## Connections +- [[bu-tan-ji-shu]] ← 思想关联 ← [[yang-xia-ri-ji-5]](乔布斯.skill是bu-tan-ji-shu的内容来源) +- [[养龙虾5天血泪史]] ← 工具 ← [[yang-xia-ri-ji-5]](都用OpenClaw作为运行平台) + +## Contradictions +- 与"AI应该像人一样通用"思维冲突:本文主张蒸馏特定人格形成专一领域的认知操作系统 diff --git a/wiki/sources/yi-yu-dian-xing-meng-zhong-ren.md b/wiki/sources/yi-yu-dian-xing-meng-zhong-ren.md new file mode 100644 index 00000000..5a891360 --- /dev/null +++ b/wiki/sources/yi-yu-dian-xing-meng-zhong-ren.md @@ -0,0 +1,69 @@ +--- +title: 一语点醒梦中人 +type: source +tags: [哲学, 智慧, 古典文学] +date: 2025-12-18 +source_file: ../raw/AI/一语点醒梦中人.md +--- + +## Summary +- 核心主题:中国古典哲学智慧与人生修养 +- 问题域:如何在人生困境中保持内心平静与智慧 +- 方法/机制:收录王维、庄子、老子等古人名言,附深度解读 +- 结论/价值:为现代人提供东方哲学智慧参考 + +## Key Claims + +## Key Quotes + +### 行到水穷处,坐看云起时 +> "行到水穷处,坐看云起时" — 王维 +- 王维被誉为"诗佛",其诗作充满禅意与佛学智慧 +- "行到水穷处"象征人生的尽头和困境 +- "坐看云起时"表达了在困境中放下执着,静观云起,顿悟人生真谛 + +### 执一守中,有劳而作,言行意合,自然而行 +- 融合儒家"执两用中"与道家"守中"思想 +- 以"中"为天下根本,避免极端,在动态平衡中守持正道 +- 有劳而作:以积极行动契合天道,在劳作中体悟自然之理 + +### 唯忘机可以消众机,唯懵懂可以祓不吉祥 +> 此句出自清代重臣**曾国藩**的《治心经·诚心篇》 +- 忘机:摒弃心机智巧,保持淳朴自然的心态 +- 以无争无求、大智若愚的态度应对复杂环境 + +### 大智若愚,大巧若拙 +> 出自《老子·第四十五章》:「大直若屈,大巧若拙,大辩若讷。」 +- 真正的智慧看似愚钝,真正的灵巧看似笨拙 +- 高人往往不露锋芒,以质朴掩藏才智 + +### 和光同尘 +> 出自《老子·第五十六章》:「和其光,同其尘。」 +- 收敛光芒,混同尘俗 +- 不标新立异,与世无争以保全自身 + +### 飘风不终朝,骤雨不终日 +> 出自《老子·第二十三章》 +- 狂风暴雨不会持续整天,喻困境终会过去 +- 与西方谚语"This too shall pass"异曲同工 + +### 一切有为法,如梦幻泡影,如露亦如电,应作如是观 +> 出自《金刚经》最后一句 +- 世间一切因缘和合的现象(有为法),皆如梦境、泡沫、露水、闪电般虚幻短暂 +- 教导以"空性"智慧观照,不执着于表象 + +### 知其不可奈何而安之若命 +> 出自《庄子·内篇·人间世》:"知其不可奈何而安之若命,德之至也。" +- 知道有些事情是自己无法改变的安然接受 +- 在尽了全部努力后,对于依然无法改变的局面,保持内心平静 + +## Key Concepts +- [[佛学智慧]]:王维晚年转向佛教,形成空寂、淡泊的心境 +- [[中庸之道]]:儒家"执两用中",避免极端,动态平衡 +- [[无为而治]]:道家"守中"、"和光同尘"的处世哲学 +- [[大智若愚]]:藏锋守拙的生存哲学 +- [[东方哲学]]:绝处逢生的智慧,面对困境的从容态度 + +## Connections + +## Contradictions diff --git a/wiki/sources/yong-docker-an-zhuang-apache-superset.md b/wiki/sources/yong-docker-an-zhuang-apache-superset.md new file mode 100644 index 00000000..125e2b7c --- /dev/null +++ b/wiki/sources/yong-docker-an-zhuang-apache-superset.md @@ -0,0 +1,25 @@ +--- +title: "用Docker安装Apache Superset" +type: source +tags: [Docker, Apache Superset, 数据可视化] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠用Docker安装Apache Superset.md +--- + +## Summary +- 核心主题:Apache Superset数据可视化平台Docker部署 +- 问题域:BI仪表盘和数据探索 +- 结论/价值:Docker一键部署自托管数据可视化平台 + +## Key Concepts +- [[数据可视化]]:BI仪表盘平台 +- [[Docker部署]]:容器化自托管方案 + +## Key Entities +- [[Apache-Superset]]:开源数据可视化平台 + +## Connections +- [[Docker]] ← 容器 ← [[Apache-Superset安装]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/yong-docker-an-zhuang-homarr.md b/wiki/sources/yong-docker-an-zhuang-homarr.md new file mode 100644 index 00000000..531bb879 --- /dev/null +++ b/wiki/sources/yong-docker-an-zhuang-homarr.md @@ -0,0 +1,29 @@ +--- +title: "用Docker安装Homarr" +type: source +tags: [Docker, Homarr, 仪表盘] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠用Docker安装Homarr.md +--- + +## Summary +- 核心主题:Homarr服务器仪表盘Docker部署 +- 问题域:服务器状态聚合展示 +- 结论/价值:docker-compose快速部署Homarr仪表盘 + +## Key Claims +- 使用ghcr.io/homarr-labs/homarr镜像 +- 端口7575映射 +- 支持ALL_PROXY环境变量配置代理 + +## Key Concepts +- [[服务器仪表盘]]:聚合展示服务器状态 + +## Key Entities +- [[Homarr]]:服务器状态仪表盘工具 + +## Connections +- [[Docker]] ← 容器 ← [[Homarr安装]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/yong-docker-an-zhuang-it-tools.md b/wiki/sources/yong-docker-an-zhuang-it-tools.md new file mode 100644 index 00000000..f9a81a58 --- /dev/null +++ b/wiki/sources/yong-docker-an-zhuang-it-tools.md @@ -0,0 +1,25 @@ +--- +title: "用Docker安装it-tools" +type: source +tags: [docker, it-tools] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠用Docker安装it-tools.md +--- + +## Summary +- 核心主题:it-tools工具集通过Docker部署 +- 问题域:IT运维常用工具的一站式部署 +- 方法/机制:使用corentinth/it-tools镜像,配置内存限制128M +- 结论/价值:提供URL编码/解码、JSON格式化等常用IT工具的Web界面 + +## Key Claims +- it-tools是面向IT从业者的工具集合 +- 部署时配置stdin_open和tty以支持交互模式 +- 建议内存限制为128M以控制资源占用 + +## Key Concepts +- [[it-tools]]:IT从业者常用工具Web集合 +- [[Docker容器化部署]]:轻量级工具容器部署 + +## Connections +- [[it-tools]] ← 部署方式 ← [[Docker容器化部署]] diff --git a/wiki/sources/yong-docker-an-zhuang-jellyfin.md b/wiki/sources/yong-docker-an-zhuang-jellyfin.md new file mode 100644 index 00000000..f05f8ca2 --- /dev/null +++ b/wiki/sources/yong-docker-an-zhuang-jellyfin.md @@ -0,0 +1,31 @@ +--- +title: "用Docker安装Jellyfin" +type: source +tags: [docker, jellyfin, synology, nas] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠用Docker安装Jellyfin.md +--- + +## Summary +- 核心主题:Jellyfin媒体服务器通过Docker在群晖NAS上的安装配置 +- 问题域:家庭媒体中心搭建,支持电影和剧集管理及硬件转码 +- 方法/机制:使用nyanmisaka/jellyfin镜像,通过Docker Compose部署,挂载硬件渲染设备启用Intel QuickSync转码 +- 结论/价值:提供开源自托管媒体服务器解决方案,替代Plex/Emby + +## Key Claims +- 群晖建议使用特定UID:GID(1026:100)运行容器以确保文件权限正确 +- 通过挂载/dev/dri设备可启用Intel QuickSync硬件转码,显著降低CPU负载 +- Jellyfin访问URL通过JELLYFIN_PublishedServerUrl环境变量配置 + +## Key Concepts +- [[Jellyfin]]:开源媒体服务器,支持电影/剧集管理和转码 +- [[Docker容器化部署]]:通过Docker Compose在NAS上部署服务 +- [[硬件转码]]:Intel QuickSync通过GPU加速视频转码 + +## Key Entities +- [[Synology NAS]]:群晖NAS,Jellyfin的运行宿主 +- [[nyanmisaka/jellyfin]]:针对中文环境优化的Jellyfin Docker镜像 + +## Connections +- [[Synology NAS + Xiaoya Alist + CloudDrive2 + Plex 家庭影视平台]] ← 相关 → [[用Docker安装Jellyfin]] +- [[Jellyfin]] ← 部署方式 ← [[Docker容器化部署]] diff --git a/wiki/sources/yong-docker-an-zhuang-portainer.md b/wiki/sources/yong-docker-an-zhuang-portainer.md new file mode 100644 index 00000000..05039363 --- /dev/null +++ b/wiki/sources/yong-docker-an-zhuang-portainer.md @@ -0,0 +1,29 @@ +--- +title: "用Docker安装Portainer" +type: source +tags: [docker, portainer] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠用Docker安装Portainer.md +--- + +## Summary +- 核心主题:Portainer容器管理平台的Docker安装配置 +- 问题域:Docker可视化运维管理 +- 方法/机制:使用portainer/portainer-ce:lts镜像,通过Docker Compose部署 +- 结论/价值:提供Web界面管理Docker容器、卷和网络 + +## Key Claims +- Portainer CE LTS版本提供长期支持 +- 需要挂载/var/run/docker.sock以访问宿主Docker守护进程 +- 默认使用9443端口提供HTTPS管理界面 + +## Key Concepts +- [[Portainer]]:Docker可视化容器管理平台 +- [[Docker容器化部署]]:通过Docker Compose部署 + +## Key Entities +- [[Portainer]]:容器管理平台 + +## Connections +- [[Portainer]] ← 部署方式 ← [[Docker容器化部署]] +- [[用Docker安装Jellyfin]] ← 共用模式 ← [[用Docker安装Portainer]] diff --git a/wiki/sources/yong-docker-an-zhuang-transmission.md b/wiki/sources/yong-docker-an-zhuang-transmission.md new file mode 100644 index 00000000..ba966181 --- /dev/null +++ b/wiki/sources/yong-docker-an-zhuang-transmission.md @@ -0,0 +1,28 @@ +--- +title: "用Docker安装transmission" +type: source +tags: [docker, transmission, bt下载] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠用Docker安装transmission.md +--- + +## Summary +- 核心主题:Transmission BT下载客户端通过Docker部署 +- 问题域:家庭下载服务器搭建 +- 方法/机制:使用linuxserver/transmission镜像,配置Web UI访问和Peer端口 +- 结论/价值:提供远程Web界面管理BT下载任务 + +## Key Claims +- Web UI端口9091,Peer监听端口51413(TCP/UDP) +- 支持通过USER/PASS环境变量配置Web UI认证 +- 配置文件挂载到宿主机/data目录,下载目录挂载到/downloads + +## Key Concepts +- [[Transmission]]:开源BT下载客户端 +- [[Docker容器化部署]]:下载服务容器化部署 + +## Key Entities +- [[Transmission]]:BT下载客户端 + +## Connections +- [[Transmission]] ← 部署方式 ← [[Docker容器化部署]] diff --git a/wiki/sources/yong-docker-zhong-an-zhuang-navidrome.md b/wiki/sources/yong-docker-zhong-an-zhuang-navidrome.md new file mode 100644 index 00000000..03a930a2 --- /dev/null +++ b/wiki/sources/yong-docker-zhong-an-zhuang-navidrome.md @@ -0,0 +1,29 @@ +--- +title: "用Docker安装Navidrome" +type: source +tags: [Docker, Navidrome, 音乐] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠用Docker中安装Navidrome.md +--- + +## Summary +- 核心主题:Navidrome音乐服务器Docker部署 +- 问题域:自托管音乐流媒体服务 +- 结论/价值:docker-compose快速部署可道音乐服务器 + +## Key Claims +- 使用deluan/navidrome:latest镜像 +- 配置ND_LOGLEVEL、ND_ENABLETRANSCODINGCONFIG等环境变量 +- 支持自动转码和下载 + +## Key Concepts +- [[音乐流媒体]]:自托管音乐服务器 + +## Key Entities +- [[Navidrome]]:开源音乐流媒体服务器 + +## Connections +- [[Docker]] ← 容器 ← [[Navidrome安装]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/youtube-channel-id.md b/wiki/sources/youtube-channel-id.md new file mode 100644 index 00000000..a6e0aa4b --- /dev/null +++ b/wiki/sources/youtube-channel-id.md @@ -0,0 +1,26 @@ +--- +title: "How to get Youtube Channel ID" +type: source +tags: [YouTube, n8n, Workflow] +date: 2025-03-16 +source_file: raw/Technical/How to get Youtube Channel ID.md +--- + +## Summary +- 核心主题:获取YouTube频道ID的方法 +- 方法:通过view-source页面查询channel_id参数 +- 用途:用于n8n工作流集成 + +## Key Claims +- 在YouTube频道页面,通过view-source可以找到channel_id +- Channel ID格式:`UCPlwvN0w4qFSP1FllALB92w` + +## Key Entities +- [[YouTube]]:视频平台 +- [[n8n]]:工作流自动化工具 + +## Connections +- [[n8n]] ← 集成 ← [[YouTube-Channel-ID]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/youtube-content-pipeline.md b/wiki/sources/youtube-content-pipeline.md new file mode 100644 index 00000000..c69de3b5 --- /dev/null +++ b/wiki/sources/youtube-content-pipeline.md @@ -0,0 +1,30 @@ +--- +title: "YouTube Content Pipeline" +type: source +tags: [OpenClaw, YouTube, Content Pipeline, Telegram, Asana] +date: 2026-04-14 +--- + +## Summary +- 核心主题:YouTube 内容管道 +- 问题域:每日 YouTube 创作者如何自动化灵感发现 +- 方法/机制:每小时 cron 扫描 AI 新闻 + 90 天视频目录 + 向量去重 +- 结论/价值:不再重复同一话题 + +## Key Claims +- 每小时 cron 扫描 AI 新闻(网页 + X/Twitter),向 Telegram 推送创意 +- 维护 90 天视频目录(含观看量/话题分析) +- SQLite + 向量嵌入存储所有创意,语义去重 +- Slack 分享链接时自动研究并创建 Asana 任务卡 + +## Key Concepts +- [[YouTube Content Pipeline]]:YouTube 内容管道 +- [[Semantic Dedup]]:语义去重 +- [[Content Scout]]:内容侦察 + +## Connections +- [[OpenClaw]] ← 运行 ← [[YouTube Content Pipeline]] +- [[Telegram]] ← 推送 ← [[YouTube Content Pipeline]] + +## Contradictions +- 尚未发现 diff --git a/wiki/sources/youtube-rss-feed.md b/wiki/sources/youtube-rss-feed.md new file mode 100644 index 00000000..72051c8b --- /dev/null +++ b/wiki/sources/youtube-rss-feed.md @@ -0,0 +1,47 @@ +--- +title: "How to Get the RSS Feed For Any YouTube Channel" +type: source +tags: [youtube, rss, tutorial] +sources: ["https://chuck.is/yt-rss/"] +last_updated: 2025-10-10 +--- + +## Summary + +- 核心主题:获取任意YouTube频道RSS订阅源的方法 +- 问题域:YouTube移除了原生RSS订阅按钮后如何获取频道RSS +- 方法/机制:通过查看页面源代码搜索channel_id构建RSS Feed URL +- 结论/价值:提供免费、简单的方法获取YouTube频道RSS,无需第三方服务 + +## Key Claims + +- YouTube移除了RSS订阅按钮因为这会让用户无需访问网站就获取内容,影响YouTube收益 +- 通过View Page Source搜索"channel_id="即可获取对应频道的RSS Feed URL +- RSS Feed URL格式:https://www.youtube.com/feeds/videos.xml?channel_id=UCHkYOD-3fZbuGhwsADBd9ZQ + +## Key Quotes + +> "I decided to share this because doing a quick search yielded terrible results (you should NOT be signing up for some service in order to get a YouTube account's RSS feed!)" — 作者动机 + +## Key Concepts + +- [[RSS Feed]]:Really Simple Syndication,用于聚合订阅内容的数据格式 +- [[YouTube Channel ID]]:YouTube频道唯一标识符 +- [[channel_id]]:URL参数,用于构建YouTube RSS订阅源 + +## Key Entities + +- [[YouTube]]:视频平台,移除了原生RSS按钮 +- [[Chuck Carroll]]:文章作者 + +## Connections + +- [[YouTube RSS]] ← 工具 ← [[RSS阅读器]] +- [[YouTube]] ← 获取方式 ← [[查看页面源代码]] + +## Contradictions + +- 与付费服务冲突: + - 冲突点:是否需要付费服务获取YouTube RSS + - 当前观点:通过查看源代码免费获取 + - 对方观点:需要注册第三方服务 diff --git a/wiki/sources/zai-synology-nas-shang-an-zhuang-clouddrive2.md b/wiki/sources/zai-synology-nas-shang-an-zhuang-clouddrive2.md new file mode 100644 index 00000000..438db1c3 --- /dev/null +++ b/wiki/sources/zai-synology-nas-shang-an-zhuang-clouddrive2.md @@ -0,0 +1,33 @@ +--- +title: "在Synology NAS上安装CloudDrive2" +type: source +tags: [synology, nas, clouddrive2] +date: 2026-04-14 +source_file: raw/Technical/Home Office/🟠在Synology NAS上安装CloudDrive2.md +--- + +## Summary +- 核心主题:在Synology NAS上安装CloudDrive2云盘同步应用 +- 问题域:NAS应用安装与配置 +- 方法/机制:通过套件中心添加矿神源后安装 +- 结论/价值:实现阿里云盘等云存储与NAS的同步 + +## Key Claims +- DSM 7+版本需要在root下执行一条命令修复权限 +- 安装后通过Web界面配置,需要使用阿里云盘app扫描二维码授权 +- 建议仅授权资源目录而非备份目录 + +## Key Concepts +- [[CloudDrive2]]:云盘同步应用,支持多种云存储服务 +- [[矿神源]]:Synology第三方套件源 + +## Key Entities +- [[Synology NAS]]:网络附加存储设备 +- [[CloudDrive2]]:云盘同步工具 +- [[阿里云盘]]:阿里巴巴云存储服务 + +## Connections +- [[Synology NAS]] ← 安装 ← [[CloudDrive2]] +- [[CloudDrive2]] ← 授权 ← [[阿里云盘]] + +## Contradictions diff --git a/wiki/sources/zai-ubuntu-an-zhuang-ollama-bing-yun-hang-qwen2-5-coder-7b.md b/wiki/sources/zai-ubuntu-an-zhuang-ollama-bing-yun-hang-qwen2-5-coder-7b.md new file mode 100644 index 00000000..a047d5ab --- /dev/null +++ b/wiki/sources/zai-ubuntu-an-zhuang-ollama-bing-yun-hang-qwen2-5-coder-7b.md @@ -0,0 +1,93 @@ +--- +title: "在 Ubuntu 安装 Ollama 并运行 Qwen2.5-Coder 7B" +type: source +tags: [ollama, qwen, qwen-coder, ubuntu, local-llm] +date: 2026-04-13 +source_file: raw/Technical/Home Office/在 Ubuntu 安装 Ollama 并运行 Qwen2.5‑Coder 7B.md +--- + +## Summary +- 核心主题:Ubuntu服务器安装Ollama运行Qwen2.5-Coder 7B本地大模型 +- 问题域:本地LLM部署、模型服务化、API开放 +- 方法/机制:Ollama一键安装 + systemd服务 + REST API暴露 +- 结论/价值:构建本地AI编码助手基础设施 + +## Key Claims +- Ollama提供一键式LLM部署 +- Qwen2.5-Coder 7B适合代码生成和DevOps任务 +- 默认只监听127.0.0.1,需配置才能远程访问 +- GPU加速开箱即用(需CUDA) + +## Key Concepts +- [[本地LLM]]:本地部署的大语言模型 +- [[Ollama]]:LLM运行平台 +- [[Qwen2.5-Coder]]:阿里通义千问代码模型 +- [[REST API]]:模型服务接口 + +## Key Entities +- [[Ollama]]:LLM服务平台 +- [[Qwen2.5-Coder]]:代码生成模型 +- [[Ubuntu]]:部署系统 + +## 系统要求 +| 资源 | 最低 | 推荐 | +|------|------|------| +| CPU | 4核 | 8+核 | +| RAM | 8GB | 16GB | +| GPU | 无需 | NVIDIA | +| 磁盘 | 10GB | 20GB | + +模型大小:约4.5GB + +## 安装步骤 +```bash +# 1. 安装 +curl -fsSL https://ollama.com/install.sh | sh + +# 2. 下载模型 +ollama pull qwen2.5-coder:7b + +# 3. 运行 +ollama run qwen2.5-coder:7b +``` + +## API调用 + +### REST API +```bash +curl http://localhost:11434/api/chat -d '{ + "model": "qwen2.5-coder:7b", + "messages": [{"role": "user", "content": "Write a bash script"}] +}' +``` + +### Python调用 +```python +from ollama import chat +response = chat(model="qwen2.5-coder:7b", messages=[{"role": "user", "content": "..."}]) +``` + +### NodeJS调用 +```javascript +import ollama from 'ollama' +const response = await ollama.chat({ model: 'qwen2.5-coder:7b', messages: [...] }) +``` + +## 开放远程API +编辑`/etc/systemd/system/ollama.service`,添加: +```ini +Environment="OLLAMA_HOST=0.0.0.0" +``` +然后: +```bash +sudo systemctl daemon-reload +sudo systemctl restart ollama +``` + +## 推荐搭配 +| 工具 | 用途 | +|------|------| +| Open WebUI | ChatGPT UI | +| n8n | AI自动化 | +| OpenClaw | AI coding agent | +| LangChain | Agent framework | \ No newline at end of file diff --git a/wiki/sources/zi-sheng-cheng-shi-pin-wang-zhan-tui-jian.md b/wiki/sources/zi-sheng-cheng-shi-pin-wang-zhan-tui-jian.md new file mode 100644 index 00000000..0693cc34 --- /dev/null +++ b/wiki/sources/zi-sheng-cheng-shi-pin-wang-zhan-tui-jian.md @@ -0,0 +1,29 @@ +--- +title: "文字生成视频网站推荐" +type: source +tags: [text-to-video, ai-tools, video-generation] +date: 2025-12-18 +source_file: raw/AI/文字生成视频网站推荐.md +--- + +## Summary +- 核心主题:文字生成视频AI工具推荐 +- 问题域:AI视频生成工具选型 +- 方法/机制:对比分析多款工具的功能、价格、适用场景 +- 结论/价值:性价比排序和场景推荐 + +## Key Claims +- 万彩AI:完全免费且功能全面,适合预算有限用户 +- 百度AI开放平台:免费套餐+多模态技术,适合技术型用户 +- Zeemo:高精度字幕+多语言支持,适合全球化内容 +- Vizard:免费版基础功能,适合长视频处理 + +## Key Entities +- [[万彩AI]]:免费文字生成视频工具 +- [[百度AI开放平台]]:大厂多模态技术背书 +- [[Zeemo]]:蓝色脉动公司,95种语言字幕 +- [[Vizard]]:视频自动剪辑工具 + +## Connections +- [[万彩AI]] ← 免费方案 ← [[文字生成视频]] +- [[百度AI开放平台]] ← 大厂方案 ← [[文字生成视频]] \ No newline at end of file diff --git a/wiki/sources/使用Claude自动生成N8N工作流的实操教程.md b/wiki/sources/使用Claude自动生成N8N工作流的实操教程.md new file mode 100644 index 00000000..83d6cafe --- /dev/null +++ b/wiki/sources/使用Claude自动生成N8N工作流的实操教程.md @@ -0,0 +1,41 @@ +--- +title: "使用Claude自动生成N8N工作流的实操教程" +type: source +tags: [n8n, claude, workflow, MCP] +sources: [] +last_updated: 2026-04-14 +source_file: raw/Technical/Workflow/使用Claude自动生成N8N工作流的实操教程.md +--- + +## Summary +- 核心主题:通过 Claude + n8n-mcp 实现自然语言生成 n8n 工作流的完整实操 +- 问题域:N8N 初学者不知如何设计工作流架构和选择节点 +- 方法/机制:安装 Node.js、配置 n8n-mcp、导入 Prompt,让 Claude 自动选节点和写代码 +- 结论/价值:Claude 能完成 80%-90% 的工作流布局,剩余细节需人工修正 + +## Key Claims +- n8n-mcp 提供 543 个 n8n 节点访问,包含 271 个 AI 能力节点文档 +- 环境搭建需要 Node.js 和 Claude 桌面端 +- API Key 配置错误会导致连接失败 +- Claude 生成的脚本约有 10%-20% 错误率 +- 使用 Opensea 模型和 extended thinking 模式效果更佳 + +## Key Quotes +> "本视频系统地介绍了利用 Claude 智能助手自动生成 N8N 工作流的完整流程,从环境搭建、关键配置到提示词导入、实际任务执行和调试改进。" + +## Key Concepts +- [[n8n-mcp]]:n8n 多功能控制面板,连接 Claude 与 n8n +- [[自然语言工作流生成]]:输入自然语言描述让 AI 自动生成工作流 +- [[Extended Thinking]]:Claude 深度推理模式,提升代码生成质量 +- [[Opensea模型]]:Claude 子模型,专为代码生成优化 + +## Key Entities +- [[n8n]]:开源工作流自动化工具 +- [[Claude]]:AI 助手,可读取指令自动生成代码 +- [[Node.js]]:运行环境 +- [[Google Sheets]]:集成目标,用于新闻爬取案例 + +## Connections +- [[Claude]] ← 连接 ← [[n8n]] ← via ← [[n8n-mcp]] +- [[自然语言工作流生成]] ← 使用 ← [[Opensea模型]] +- [[自然语言工作流生成]] ← 使用 ← [[Extended Thinking]] \ No newline at end of file