Auto-sync: 2026-04-20 07:08
This commit is contained in:
@@ -2,9 +2,9 @@
|
||||
title: "Cloud Operating Model: Key Strategies and Best Practices"
|
||||
type: source
|
||||
tags: [cloud-operating-model, cloud-governance, finops]
|
||||
date: 2025-02-07
|
||||
date: 2025-03-01
|
||||
sources: []
|
||||
last_updated: 2025-02-07
|
||||
last_updated: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
|
||||
@@ -0,0 +1,48 @@
|
||||
---
|
||||
title: "Learning Sessions Cloud Transformation Programme-Deploying RDS via Terraform"
|
||||
type: source
|
||||
tags: [Terraform, RDS, IaC, CTP]
|
||||
date: 2026-04-14
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Cloud & DevOps/Public-Cloud-Learning-Sessions/03_Terraform/learning-sessions-cloud-transformation-programme-deploying-rds-via-terraform.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:通过 Terraform 部署 Amazon RDS,推广基础设施即代码(IaC)方法
|
||||
- 问题域:RDS 部署方式选择(控制台 vs IaC)、模块化基础设施管理
|
||||
- 方法/机制:使用 Terragrunt(Terraform 包装器)进行模块化部署,SRE 核心模块和 Gruntwork 模块
|
||||
- 结论/价值:IaC 提供速度、灵活性、一致性、灾难恢复、文档化和自动化优势
|
||||
|
||||
## Key Claims
|
||||
- IaC 相比控制台部署更适合任何规模的 RDS — 代码即文档
|
||||
- 推荐使用 Gruntwork RDS 服务而非裸机 RDS 模块(预建 KMS 加密和 CloudWatch 告警功能)
|
||||
- SRE 核心模块功能不如 Gruntwork 服务完善
|
||||
- 使用 Terragrunt 保持代码整洁,避免变量重复
|
||||
- 生产环境应使用标记版本而非 master 分支以保证稳定性
|
||||
|
||||
## Key Quotes
|
||||
> "We use Terragrunt, which is basically it's a wrapper around Terraform, and it allows you to keep your code clean and you're not repeating your variables all the time." — Greg, DBRE Team
|
||||
|
||||
## Key Concepts
|
||||
- [[IaC]]:基础设施即代码,通过声明式配置管理云资源
|
||||
- [[Terragrunt]]:Terraform 的包装工具,提供模块化、变量共享和环境隔离
|
||||
- [[RDS]]:Amazon 关系数据库服务
|
||||
- [[CloudWatch]]:AWS 云监控服务,用于仪表板和告警
|
||||
- [[KMS]]:AWS 密钥管理服务,用于数据加密
|
||||
|
||||
## Key Entities
|
||||
- [[Greg]]:DBRE 团队成员,演讲者
|
||||
- [[Gruntwork]]:提供预建基础设施模块的公司
|
||||
- [[AWS]]:云服务提供商
|
||||
- [[Cloud Transformation Programme]]:云转型项目(CTP)
|
||||
|
||||
## Connections
|
||||
- [[Terragrunt]] ← uses ← [[Terraform]]
|
||||
- [[RDS]] ← deployed_by ← [[Terragrunt]]
|
||||
- [[RDS]] ← monitored_by ← [[CloudWatch]]
|
||||
- [[RDS]] ← encrypted_by ← [[KMS]]
|
||||
- [[Gruntwork]] ← provides ← [[RDS-Module]]
|
||||
|
||||
## Contradictions
|
||||
- 无
|
||||
63
wiki/sources/agency-agents-readme.md
Normal file
63
wiki/sources/agency-agents-readme.md
Normal file
@@ -0,0 +1,63 @@
|
||||
---
|
||||
title: "The Agency: AI Specialists Ready to Transform Your Workflow"
|
||||
type: source
|
||||
tags: [ai-agent, open-source, multi-agent]
|
||||
date: 2025-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/README.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:The Agency 开源 AI 智能体集合项目,提供 144+ 个跨 12 个部门的专业化 AI Agent
|
||||
- 问题域:AI Agent 缺乏标准化设计和多工具集成方案
|
||||
- 方法/机制:通过预定义 Agent 模板(身份、使命、交付物、工作流、成功指标)实现专业化,每个 Agent 具备鲜明个性、明确交付物、验证工作流和学习记忆
|
||||
- 结论/价值:提供可直接使用的多工具集成方案,覆盖开发、设计、媒体、销售、营销、产品、项目管理、测试、支持、空间计算、游戏开发、学术研究等领域
|
||||
|
||||
## Key Claims
|
||||
- The Agency 提供 144+ 专业 AI Agent,跨越 12 个部门
|
||||
- 每个 Agent 具备鲜明个性、明确交付物、成功指标、经过验证的工作流、学习记忆和真实场景测试
|
||||
- 支持 Claude Code、GitHub Copilot、Cursor、Windsurf、OpenCode、OpenClaw 等 11 种 AI 工具
|
||||
- Agent 设计遵循五大原则:强个性、清晰交付物、成功指标、验证工作流、学习记忆
|
||||
|
||||
## Key Quotes
|
||||
> "Think of it as: Assembling your dream team, except they're AI specialists who never sleep, never complain, and always deliver." — 项目定位
|
||||
|
||||
> "I don't just test your code - I default to finding 3-5 issues and require visual proof for everything." — Evidence Collector
|
||||
|
||||
> "You're not marketing on Reddit - you're becoming a valued community member who happens to represent a brand." — Reddit Community Builder
|
||||
|
||||
## Key Concepts
|
||||
- [[智能体设计规范]]:Agent 文件结构、身份与记忆、核心使命、技术交付物、工作流程的完整定义
|
||||
- [[智能体设计原则]]:五大设计原则(鲜明性格、明确交付物、成功指标、经过验证的工作流、学习记忆)
|
||||
- [[Multi-Agent Team]]:多 Agent 协作架构,每个 Agent 有独立角色、人格、优化的模型
|
||||
- [[MCP Builder]]:Model Context Protocol 服务器构建者,扩展 AI Agent 能力
|
||||
|
||||
## Key Entities
|
||||
- [[msitarzewski]]:GitHub 项目作者,The Agency 发起者
|
||||
- [[Claude Code]]:原生支持 .md 格式 Agent 的 AI 编程工具
|
||||
- [[Cursor]]:通过 .cursor/rules 集成 Agent 的 IDE
|
||||
- [[Windsurf]]:通过 .windsurfrules 集成 Agent 的 AI IDE
|
||||
|
||||
## Connections
|
||||
- [[AI智能体]] ← extends ← The Agency
|
||||
- [[OpenClaw]] ← integrates_with ← The Agency(通过 SOUL.md + AGENTS.md + IDENTITY.md)
|
||||
- [[Multi-Agent Team]] ← uses_pattern ← The Agency(层级结构、共识投票等架构)
|
||||
|
||||
## Contradictions
|
||||
- 无明显冲突
|
||||
|
||||
## Agent 部门分类
|
||||
- Engineering Division(24 个):前端、后端、移动端、AI、DevOps、安全等
|
||||
- Design Division(7 个):UI、UX、品牌、图像提示词等
|
||||
- Paid Media Division(7 个):PPC、搜索分析、广告审计等
|
||||
- Sales Division(8 个):外呼、发现、交易策略、销售工程等
|
||||
- Marketing Division(22 个):增长、内容、社交媒体、SEO 等
|
||||
- Product Division(5 个): sprint 优先级、趋势研究、反馈综合等
|
||||
- Project Management Division(6 个):制片、项目协调、实验追踪等
|
||||
- Testing Division(8 个):证据收集、现实检查、性能基准等
|
||||
- Support Division(6 个):支持响应、分析、财务等
|
||||
- Spatial Computing Division(6 个):XR、visionOS、终端集成等
|
||||
- Specialized Division(26 个):MCP 构建者、合规审计、销售自动化等
|
||||
- Game Development Division(17 个):跨引擎、Unity、Unreal、Godot、Blender、Roblox
|
||||
- Academic Division(5 个):人类学家、地理学家、历史学家等
|
||||
42
wiki/sources/baoyu-skills.md
Normal file
42
wiki/sources/baoyu-skills.md
Normal file
@@ -0,0 +1,42 @@
|
||||
---
|
||||
title: "JimLiu/baoyu-skills"
|
||||
type: source
|
||||
tags: []
|
||||
date: 2026-04-19
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Skills/baoyu-skills.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:Claude Code 技能集,提供内容生成、AI 图像生成、日常效率工具
|
||||
- 问题域:AI 辅助内容创作、多平台发布、图像生成、文档转换
|
||||
- 方法/机制:基于 Claude Code 的 Skill 系统,通过 npx 命令调用各类技能
|
||||
- 结论/价值:为 Claude Code 用户提供开箱即用的 AI 创作和发布能力
|
||||
|
||||
## Key Claims
|
||||
- baoyu-skills 是宝玉分享的 Claude Code 技能集,可通过 `npx skills add jimliu/baoyu-skills` 快速安装
|
||||
- 技能分为内容技能、AI 生成技能、工具技能三大类,共 17+ 个技能
|
||||
- 支持多平台发布(X、微信、微博),支持多种图像生成服务商
|
||||
|
||||
## Key Quotes
|
||||
> "宝玉分享的 Claude Code 技能集,提升日常工作效率。" — 项目描述
|
||||
|
||||
## Key Concepts
|
||||
- [[Claude Code Skills]]:写给 Claude 的"说明书",将反复执行的任务拆解为 AI 可稳定复用的流程
|
||||
- [[AI 图像生成]]:通过 baoyu-imagine 支持多种服务商(OpenAI、Google、MiniMax、Replicate 等)
|
||||
- [[内容发布自动化]]:通过 baoyu-post-to-x、baoyu-post-to-wechat、baoyu-post-to-weibo 实现多平台发布
|
||||
|
||||
## Key Entities
|
||||
- [[JimLiu]]:baoyu-skills 项目作者,宝玉
|
||||
- [[Claude Code]]:Anthropic 提供的 AI 编程 CLI 工具
|
||||
- [[ClawHub]]:OpenClaw 的插件市场
|
||||
|
||||
## Connections
|
||||
- [[baoyu-imagine]] ← generated_by ← [[Claude Code Skills]]
|
||||
- [[Claude Code Skills]] ← maintained_by ← [[JimLiu]]
|
||||
- [[baoyu-post-to-wechat]] ← requires ← [[微信公众号 API]]
|
||||
- [[baoyu-xhs-images]] ← generates ← [[小红书图片]]
|
||||
|
||||
## Contradictions
|
||||
- 无
|
||||
48
wiki/sources/contributing-zh-cn.md
Normal file
48
wiki/sources/contributing-zh-cn.md
Normal file
@@ -0,0 +1,48 @@
|
||||
---
|
||||
title: "为 The Agency 贡献代码"
|
||||
type: source
|
||||
tags: []
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/CONTRIBUTING_zh-CN.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:The Agency 项目的贡献指南,涵盖智能体设计规范、PR 流程和社区参与方式
|
||||
- 问题域:开源 AI 智能体项目的贡献流程标准化
|
||||
- 方法/机制:明确的智能体模板结构、设计原则、PR 审核流程
|
||||
- 结论/价值:为 AI 智能体社区贡献提供规范化指导,降低贡献门槛,提高智能体质量
|
||||
|
||||
## Key Claims
|
||||
- 智能体设计必须遵循六大原则:鲜明性格、明确交付物、成功指标、经过验证的工作流、学习记忆、真实场景测试
|
||||
- 优秀智能体必须具备:专精领域定位、独特性格语气、具体代码示例、可量化指标、分步工作流
|
||||
- PR 审核流程包括:社区评审、迭代优化、通过审核、合并上线四个阶段
|
||||
|
||||
## Key Quotes
|
||||
> "正是有像你这样的参与者,才能让这套 AI 智能体集合变得越来越好" — 贡献开场白
|
||||
|
||||
> "避免通用型'有用助手'人设" — 智能体设计核心原则
|
||||
|
||||
> "提供真实代码,而非伪代码" — 风格指南要求
|
||||
|
||||
## Key Concepts
|
||||
- [[智能体设计规范]]:智能体文件结构、身份与记忆、核心使命、技术交付物、工作流程的完整定义
|
||||
- [[智能体设计原则]]:六大设计原则,包括鲜明性格、明确交付物、成功指标、验证工作流、学习记忆
|
||||
- [[Pull Request 流程]]:从提交前准备到 PR 审核的完整流程
|
||||
- [[风格指南]]:写作风格、格式规范、代码示例的具体要求
|
||||
|
||||
## Key Entities
|
||||
- [[The Agency]]:AI 智能体集合项目
|
||||
- [[agency-agents]]:GitHub 仓库名称
|
||||
|
||||
## Connections
|
||||
- [[智能体设计规范]] ← defines ← [[The Agency]]
|
||||
- [[Pull Request 流程]] ← enables ← [[社区贡献]]
|
||||
- [[风格指南]] ← governs ← [[代码示例]]
|
||||
|
||||
## Contradictions
|
||||
- 无明显冲突
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/CONTRIBUTING_zh-CN.md]]
|
||||
@@ -0,0 +1,57 @@
|
||||
---
|
||||
title: "CTP Topic 12 Using SES SMTP service terraform module"
|
||||
type: source
|
||||
tags:
|
||||
- AWS
|
||||
- Terraform
|
||||
- SES
|
||||
- Email
|
||||
- CTP
|
||||
date: 2026-04-14
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Cloud & DevOps/Public-Cloud-Learning-Sessions/03_Terraform/ctp-topic-12-using-ses-smtp-service-terraform-module.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:利用 AWS SES SMTP Terraform 模块实现云端邮件发送
|
||||
- 问题域:传统本地 SMTP 网关向云端迁移
|
||||
- 方法/机制:SES SMTP 端点 + VPC 终端节点 + IAM 用户凭证 + Secrets Manager
|
||||
- 结论/价值:替代传统 smtbmicrofocus.com 网关,实现安全的云端邮件发送
|
||||
|
||||
## Key Claims
|
||||
- SES 是云安全部门唯一批准的云端邮件发送方案
|
||||
- SES Terraform 模块封装了 SMTP 终端节点配置,应用程序可通过标准 SMTP 协议集成
|
||||
- VPC 终端节点确保应用与 SES 通信时不访问公网
|
||||
- IAM 用户凭证转换为 SMTP 认证信息,存储在 Secrets Manager 中
|
||||
|
||||
## Key Quotes
|
||||
> "随着业务向云端迁移,使用本地 SMTP 网关已不再高效" — Christian Deckelmann
|
||||
|
||||
> "需要手动申请脱离 SES Sandbox Mode 才能提升发送限额并允许向外部地址发信" — Filos Christolakis
|
||||
|
||||
## Key Concepts
|
||||
- [[AWS SES]]:基于云的邮件发送服务
|
||||
- [[SMTP Endpoint]]:区域性邮件传输协议终端节点
|
||||
- [[Sandbox Mode]]:SES 默认限制状态
|
||||
- [[DKIM]]:电子邮件验证标准
|
||||
- [[VPC Endpoint]]:AWS 私有网络终端节点
|
||||
- [[Secrets Manager]]:凭证管理服务
|
||||
|
||||
## Key Entities
|
||||
- [[Christian Deckelman]]:Micro Focus 云架构师,分享者
|
||||
- [[Filos Christolakis]]:SES Terraform 模块开发者
|
||||
- [[Micro Focus]]:云转型企业
|
||||
|
||||
## Connections
|
||||
- [[SES SMTP Terraform Module]] ← depends_on ← [[VPC Wrapper Module]]
|
||||
- [[SES SMTP Terraform Module]] ← depends_on ← [[Secrets Manager]]
|
||||
- [[SES]] ← extends ← [[AWS Service]]
|
||||
|
||||
## Contradictions
|
||||
- 无
|
||||
|
||||
## Aliases
|
||||
- SES: Simple Email Service
|
||||
- SMTP: Simple Mail Transfer Protocol
|
||||
- DKIM: DomainKeys Identified Mail
|
||||
47
wiki/sources/ctp-topic-16-cross-account-terraform-modules.md
Normal file
47
wiki/sources/ctp-topic-16-cross-account-terraform-modules.md
Normal file
@@ -0,0 +1,47 @@
|
||||
---
|
||||
title: "CTP Topic 16 Cross-account Terraform modules"
|
||||
type: source
|
||||
tags: [terraform, cross-account, modules, ctp, devops]
|
||||
source: []
|
||||
last_updated: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Cloud & DevOps/Public-Cloud-Learning-Sessions/03_Terraform/ctp-topic-16-cross-account-terraform-modules.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:在多账号 AWS 环境中实现和管理跨账号 Terraform 模块
|
||||
- 问题域:如何在不授予直接互访权限的情况下,在一个模块中跨多个账号创建资源
|
||||
- 方法/机制:基于 Shared Account 的中心化部署方案,通过 cross-account.json 标记文件触发 ECS Deploy Runner,Assume Role 访问目标账号
|
||||
- 结论/价值:实现安全性(集中权限控制)、自动化(自动识别模块类型)、可复用性(代码不硬编码账号角色)
|
||||
|
||||
## Key Claims
|
||||
- Cross-account Modules 通过配置多个 Provider 实现在多个 AWS 账号中同时创建或管理资源
|
||||
- Shared Account 作为信任源,通过 Assume Role 方式访问目标账号,避免直接授予互访权限
|
||||
- cross-account.json 标记文件告知 Jenkins 该模块需要调用跨账号部署逻辑
|
||||
|
||||
## Key Quotes
|
||||
> "在多账号 AWS 环境中,经常需要在一个模块内跨多个账号创建资源,但直接赋予账号间互访权限存在巨大的安全风险" — Fibos
|
||||
|
||||
> "利用托管 Jenkins 的 Shared Account 作为中转站,实现中心化的跨账号部署" — Fibos
|
||||
|
||||
## Key Concepts
|
||||
- [[Cross-account Modules]]:在一个 Terraform 模块中通过配置多个 Provider,实现在多个 AWS 账号中同时创建或管理资源的功能
|
||||
- [[Shared Account]]:整个 Landing Zone 中的核心管理账号,托管 Jenkins、镜像仓库等公共服务,并作为跨账号部署的信任源
|
||||
- [[ECS Deploy Runner]]:运行在 ECS 上的 Docker 容器,负责执行 Terraform plan 和 apply 命令
|
||||
- [[Cross-account.json]]:约定俗成的标记文件,放置在模块目录中,用于告知 Jenkins 该模块需要调用跨账号部署逻辑
|
||||
|
||||
## Key Entities
|
||||
- [[Fibos]]:本次会议讲师
|
||||
- [[Gruntwork Pipeline]]:原有的单账号部署流水线
|
||||
|
||||
## Connections
|
||||
- [[Gruntwork Pipeline Deep Dive]] ← depends_on ← [[CTP Topic 16 Cross-account Terraform modules]]
|
||||
- [[AWS Multi-account Security Best Practices]] ← informs ← [[CTP Topic 16 Cross-account Terraform modules]]
|
||||
- [[Terragrunt Advanced Configuration]] ← used_by ← [[CTP Topic 16 Cross-account Terraform modules]]
|
||||
|
||||
## Contradictions
|
||||
- 与直接授予 Workload 账号间互访权限的方案冲突:
|
||||
- 冲突点:直接授予互访权限存在 Blast Radius 安全风险
|
||||
- 当前观点:通过 Shared Account 集中控制权限,Assume Role 访问
|
||||
- 对方观点:直接授予账号间互访权限简化配置
|
||||
49
wiki/sources/design-brand-guardian.md
Normal file
49
wiki/sources/design-brand-guardian.md
Normal file
@@ -0,0 +1,49 @@
|
||||
---
|
||||
title: "Brand Guardian Agent 设计文档"
|
||||
type: source
|
||||
tags: [agent, the-agency, brand-strategy, design]
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/design/design-brand-guardian.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:品牌战略与身份保护专家智能体设计
|
||||
- 问题域:品牌一致性维护、品牌价值保护、品牌战略执行
|
||||
- 方法/机制:通过身份定义、记忆系统、核心使命、工作流程、交付物模板实现品牌战略全面管理
|
||||
- 结论/价值:创建 cohesive brand identities,确保跨所有触点的品牌一致性表达
|
||||
|
||||
## Key Claims
|
||||
- Brand Guardian 是 expert brand strategist 和 guardian,创建 cohesive brand identities 并确保跨触点一致的品牌表达
|
||||
- 品牌战略必须在战术实施前建立全面的品牌基础
|
||||
- 品牌决策必须连接到业务目标和市场定位
|
||||
- 成功的品牌需要 95%+ 跨触点一致性
|
||||
|
||||
## Key Quotes
|
||||
> "Your brand's fiercest protector and most passionate advocate." — Brand Guardian 核心定位
|
||||
> "Establish comprehensive brand foundation before tactical implementation" — 品牌优先方法论
|
||||
|
||||
## Key Concepts
|
||||
- [[Brand Foundation Framework]]:品牌基础框架,包含 Purpose、Vision、Mission、Values、Personality
|
||||
- [[Visual Identity System]]:视觉识别系统,包含 Logo、Color Palette、Typography
|
||||
- [[Brand Voice]]:品牌声音与语调规范
|
||||
- [[Brand Protection]]:品牌保护策略(商标、法律、危机管理)
|
||||
|
||||
## Key Entities
|
||||
- [[The Agency]]:开源 AI 智能体集合项目,提供 144+ 跨 12 个部门的专业化 AI Agent
|
||||
|
||||
## Connections
|
||||
- [[design-ux-architect]] ← same_project ← [[design-ux-researcher]]
|
||||
- [[design-whimsy-injector]] ← extends ← [[Brand Strategy]]
|
||||
|
||||
## Contradictions
|
||||
-
|
||||
|
||||
---
|
||||
**Design Elements**:
|
||||
- Identity & Memory: 角色定义、个性、记忆系统
|
||||
- Core Mission: 三大使命(品牌基础创建、品牌一致性保护、战略性品牌演进)
|
||||
- Critical Rules: 品牌优先方法论、战略性品牌思维
|
||||
- Workflow: 四步流程(品牌发现与战略→基础开发→系统创建→实施与保护)
|
||||
- Deliverables: 品牌基础框架、视觉识别系统、品牌声音与信息传递模板
|
||||
47
wiki/sources/design-image-prompt-engineer.md
Normal file
47
wiki/sources/design-image-prompt-engineer.md
Normal file
@@ -0,0 +1,47 @@
|
||||
---
|
||||
title: "Image Prompt Engineer Agent"
|
||||
type: source
|
||||
tags: [agent, prompt-engineering, ai-image-generation, the-agency]
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/design/design-image-prompt-engineer.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:AI 图像生成提示词工程专家智能体
|
||||
- 问题域:如何将视觉概念转化为精确的提示词语言,生成专业级 AI 摄影作品
|
||||
- 方法/机制:通过分层提示词结构(主体、环境、灯光、技术、风格)实现视觉概念翻译
|
||||
- 结论/价值:为 AI 图像生成工具提供专业摄影级别的提示词设计能力
|
||||
|
||||
## Key Claims
|
||||
- Image Prompt Engineer 应始终使用结构化提示词框架,包含主体、环境、灯光、风格和技术规格
|
||||
- 有效的提示词必须使用具体、精确的摄影术语而非模糊描述
|
||||
- 提示词需要根据不同 AI 平台(Midjourney、DALL-E、Stable Diffusion、Flux)进行语法优化
|
||||
|
||||
## Key Quotes
|
||||
> "Always structure prompts with subject, environment, lighting, style, and technical specs" — 提示词结构核心原则
|
||||
|
||||
> "Use specific, concrete terminology rather than vague descriptors" — 精确术语替代模糊描述
|
||||
|
||||
> "Master the art of translating visual concepts into precise, structured language that produces stunning, professional-quality photography" — 核心使命
|
||||
|
||||
## Key Concepts
|
||||
- [[Prompt Engineering]]:将视觉概念转化为精确提示词语言的技术
|
||||
- [[Prompt Structure Framework]]:分层提示词结构框架(主体层、环境层、灯光层、技术层、风格层)
|
||||
- [[Photography Terminology]]:专业摄影术语在提示词中的应用
|
||||
- [[Platform-Specific Optimization]]:针对不同 AI 平台的提示词优化技巧
|
||||
|
||||
## Key Entities
|
||||
- [[The Agency]]:开源 AI 智能体集合项目,Image Prompt Engineer 所属项目
|
||||
- Midjourney:AI 图像生成平台,支持参数语法(--ar, --v, --style, --chaos)
|
||||
- DALL-E:OpenAI 的 AI 图像生成模型
|
||||
- Stable Diffusion:开源 AI 图像生成模型
|
||||
- Flux:开源 AI 图像生成模型
|
||||
|
||||
## Connections
|
||||
- [[Image Prompt Engineer]] ← belongs_to ← [[The Agency]]
|
||||
- [[Image Prompt Engineer]] ← depends_on ← [[Agent Design Principles]]
|
||||
- [[Image Prompt Engineer]] ← extends ← [[Agent File Structure]]
|
||||
|
||||
## Contradictions
|
||||
55
wiki/sources/design-inclusive-visuals-specialist.md
Normal file
55
wiki/sources/design-inclusive-visuals-specialist.md
Normal file
@@ -0,0 +1,55 @@
|
||||
---
|
||||
title: "Inclusive Visuals Specialist"
|
||||
type: source
|
||||
tags: [The Agency, AI Agent, Design]
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/design/design-inclusive-visuals-specialist.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:AI 图像与视频生成中的包容性视觉设计专家智能体
|
||||
- 问题域:解决基础图像/视频模型(Midjourney、Sora、Runway、DALL-E)中的系统性刻板印象和偏见问题
|
||||
- 方法/机制:通过精确的提示词工程、负向约束库、物理现实定义等技术手段,生成具有文化准确性、尊严和真实感的多元人像
|
||||
- 结论/价值:确保 AI 生成媒体以尊严、主体性和真实情境现实主义描绘各类人群,对抗默认偏见和 AI 幻觉
|
||||
|
||||
## Key Claims
|
||||
- 身份不应被视为简单的描述符输入,而是需要专业技术准确表达的领域
|
||||
- 必须明确要求不同群体中的面部结构、年龄和体型各异,防止生成"克隆面孔"
|
||||
- 必须显式负向提示任何文本、标志或生成标牌,防止 AI 发明令人反感或无意义的字符
|
||||
- 在视频生成中必须明确定义服装、头发和助行辅助工具的物理特性
|
||||
|
||||
## Key Quotes
|
||||
> "The current prompt will likely trigger the model's 'exoticism' bias. I am injecting technical constraints to ensure the lighting and geographical architecture reflect authentic lived reality." — 关键技术短语
|
||||
|
||||
> "You review AI output not just for technical fidelity, but for sociological accuracy." — 核心工作标准
|
||||
|
||||
## Key Concepts
|
||||
- [[InclusiveVisualsSpecialist]]:专注于真实人类representation的严格提示词工程师,对抗基础图像和视频模型中嵌入的系统性刻板印象
|
||||
- [[CloneFaces]]:AI 在生成多元群体时生成多个相同面孔的问题,需要通过显式约束防止
|
||||
- [[GibberishText]]:AI 在尝试非英语脚本或文化符号时发明无意义或冒犯性字符的问题
|
||||
- [[PhysicalRealityConstraints]]:在视频生成中明确定义服装、头发和助行辅助工具物理特性的技术
|
||||
- [[NegativePromptLibrary]]:针对图像和视频平台的显式负向提示库,用于阻止"AI 怪异感"
|
||||
|
||||
## Key Entities
|
||||
- [[TheAgency]]:开源 AI 智能体集合项目,汇集各类专业化 AI Agent
|
||||
- [[InclusiveVisualsSpecialist]]:The Agency 项目中的包容性视觉设计专家智能体
|
||||
- Midjourney:AI 图像生成模型
|
||||
- Sora:OpenAI 视频生成模型
|
||||
- Runway:AI 视频生成平台
|
||||
- DALL-E:OpenAI 图像生成模型
|
||||
|
||||
## Connections
|
||||
- [[InclusiveVisualsSpecialist]] ← belongs_to ← [[TheAgency]]
|
||||
- [[InclusiveVisualsSpecialist]] ← depends_on ← [[ImagePromptEngineer]]
|
||||
- [[CloneFaces]] ← counters ← [[Midjourney]]
|
||||
- [[CloneFaces]] ← counters ← [[Sora]]
|
||||
- [[GibberishText]] ← counters ← [[DALL-E]]
|
||||
- [[PhysicalRealityConstraints]] ← requires ← [[NegativePromptLibrary]]
|
||||
|
||||
## Contradictions
|
||||
- 与"批量生成多样性的过度纠正"冲突:
|
||||
- 冲突点:AI 尝试"过于多样化"时会生成符号化、不真实的构图
|
||||
- 当前观点:需要精确的技术约束来平衡真实性和多样性
|
||||
- 对方观点:简单地堆叠身份描述符即可实现多样性
|
||||
44
wiki/sources/design-ui-designer.md
Normal file
44
wiki/sources/design-ui-designer.md
Normal file
@@ -0,0 +1,44 @@
|
||||
---
|
||||
title: "UI Designer"
|
||||
type: source
|
||||
tags: [The Agency, AI Agent, UI Design]
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/design/design-ui-designer.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:The Agency 项目中的 UI Designer 智能体,专门负责创建美观、一致、可访问的用户界面
|
||||
- 问题域:视觉设计系统、组件库、像素级界面实现
|
||||
- 方法/机制:通过设计系统优先方法、可访问性合规、响应式框架、开发者协作实现
|
||||
- 结论/价值:为开发者提供完整的设计系统规范、组件样式、响应式框架和可访问性标准,实现设计到开发的高效交接
|
||||
|
||||
## Key Claims
|
||||
- UI Designer 是视觉设计系统和界面创建专家,专门负责设计系统、组件库和像素级界面实现
|
||||
- 设计系统优先方法论是核心原则,需在创建单个界面之前建立组件基础
|
||||
- 所有设计必须默认包含可访问性合规(WCAG AA 最低标准)
|
||||
- 设计交付物包含完整的组件库架构、响应式框架和开发者交接文档
|
||||
|
||||
## Key Quotes
|
||||
> "Establish component foundations before creating individual screens" — 设计系统优先方法
|
||||
> "Default requirement: Include accessibility compliance (WCAG AA minimum) in all designs" — 可访问性要求
|
||||
> "Provide clear design handoff specifications with measurements and assets" — 开发者交接标准
|
||||
|
||||
## Key Concepts
|
||||
- [[Design System]]:设计系统,包含设计令牌、组件库、响应式框架和可访问性标准
|
||||
- [[Design Token]]:设计令牌系统,用于跨平台一致性的视觉语言变量(颜色、字体、间距、阴影、过渡)
|
||||
- [[Component Library]]:组件库,包含基础组件(按钮、输入框、卡片、导航)的完整架构
|
||||
- [[Responsive Design]]:响应式设计,通过移动优先方法实现跨设备适配
|
||||
- [[WCAG AA]]:Web 内容可访问性指南 AA 级标准,4.5:1 对比度要求
|
||||
- [[Dark Theme]]:深色主题系统,通过数据属性切换实现主题切换
|
||||
|
||||
## Key Entities
|
||||
- [[The Agency]]:开源 AI 智能体集合项目,UI Designer 是其成员之一
|
||||
- [[UX Architect]]:The Agency 项目中的技术架构与 UX 专家智能体
|
||||
- [[UX Researcher]]:The Agency 项目中的用户体验研究智能体
|
||||
|
||||
## Connections
|
||||
- [[The Agency]] ← contains ← [[UI Designer]]
|
||||
- [[UX Architect]] ← collaborates_with ← [[UI Designer]]
|
||||
- [[UX Researcher]] ← provides_insights_to ← [[UI Designer]]
|
||||
56
wiki/sources/design-ux-architect.md
Normal file
56
wiki/sources/design-ux-architect.md
Normal file
@@ -0,0 +1,56 @@
|
||||
---
|
||||
title: "UX Architect Agent 设计文档"
|
||||
type: source
|
||||
tags: [The Agency, AI Agent, UX, 设计规范]
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/design/design-ux-architect.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:UX Architect(技术架构与用户体验专家)智能体设计与规范
|
||||
- 问题域:如何为开发者提供坚实的技术基础、CSS系统和清晰的实现路径
|
||||
- 方法/机制:通过设计系统变量、布局框架、组件架构、响应式策略、主题切换机制等交付物,为 LuxuryDeveloper 提供专业级 UX 基线
|
||||
- 结论/价值:消除架构决策疲劳,确保项目一致性和可扩展性,在 Premium Polish 添加前建立专业 UX 基线
|
||||
|
||||
## Key Claims
|
||||
- UX Architect 是技术架构与 UX 基础专家,为开发者创造可依赖的坚实基础
|
||||
- 默认要求所有新站点必须包含 light/dark/system 主题切换功能
|
||||
- 优先创建可扩展的 CSS 架构,建立开发者可安心构建的布局系统
|
||||
- 通过三层架构(Planner → Worker → Validator)协调 Agent 职责和技术决策
|
||||
|
||||
## Key Quotes
|
||||
> "Eliminate architectural decision fatigue for developers" — 消除开发者的架构决策疲劳
|
||||
|
||||
> "Establish professional UX baseline before premium polish is added" — 在高级优化前建立专业 UX 基线
|
||||
|
||||
> "Default requirement: Include light/dark/system theme toggle on all new sites" — 默认要求所有新站点包含主题切换
|
||||
|
||||
## Key Concepts
|
||||
- [[CSS 设计系统]]:通过 CSS 变量定义颜色、字体、间距、布局的系统化方法
|
||||
- [[响应式断点策略]]:移动优先的响应式设计方法论(320px+ → 768px+ → 1024px+ → 1280px+)
|
||||
- [[组件架构]]:分层组件系统(布局组件、内容组件、交互组件、工具组件)
|
||||
- [[主题切换]]:支持 light/dark/system 三种模式的主题管理机制
|
||||
- [[信息架构]]:页面层次结构、导航策略、内容权重的系统规划
|
||||
|
||||
## Key Entities
|
||||
- [[The Agency]]:开源 AI 智能体集合项目,汇集各类专业化 AI Agent
|
||||
- [[LuxuryDeveloper]]:负责实现 LuxuryDeveloper 级别美学的开发者 Agent
|
||||
- [[ProjectManager]]:负责任务列表管理的项目管理者 Agent
|
||||
- [[ArchitectUX]]:UX Architect Agent 的角色名称,技术架构与 UX 基础专家
|
||||
|
||||
## Connections
|
||||
- [[The Agency]] ← contains ← [[UX Architect]]
|
||||
- [[UX Architect]] → supports → [[LuxuryDeveloper]]
|
||||
- [[ProjectManager]] → provides_tasks → [[UX Architect]]
|
||||
- [[UX Architect]] ← depends_on ← [[智能体设计规范]]
|
||||
- [[UX Architect]] ← depends_on ← [[智能体设计原则]]
|
||||
|
||||
## Contradictions
|
||||
- 无
|
||||
|
||||
## Aliases
|
||||
- ArchitectUX
|
||||
- UX 架构师
|
||||
- 技术架构师
|
||||
43
wiki/sources/design-ux-researcher.md
Normal file
43
wiki/sources/design-ux-researcher.md
Normal file
@@ -0,0 +1,43 @@
|
||||
---
|
||||
title: "UX Researcher"
|
||||
type: source
|
||||
tags: [agent, the-agency, ux-research]
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/design/design-ux-researcher.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:The Agency 项目中的 UX Researcher 智能体定义,专注于用户体验研究和用户行为分析
|
||||
- 问题域:用户体验研究方法论、用户行为验证、数据驱动的设计决策
|
||||
- 方法/机制:通过定性定量研究、用户访谈、可用性测试、行为分析提供可操作洞察
|
||||
- 结论/价值:为设计团队提供基于真实用户数据的决策支持,弥合用户需求与设计解决方案之间的差距
|
||||
|
||||
## Key Claims
|
||||
- UX Researcher 通过严谨的研究方法论和数据驱动的建议,弥合用户需求与设计解决方案之间的差距
|
||||
- 研究方法优先:建立清晰的研究问题,选择适当的方法,使用适当的样本量和统计方法确保可靠洞察
|
||||
- 道德研究实践:获得适当同意,保护参与者隐私,无偏见呈现发现
|
||||
- 包容性研究:包含无障碍研究和包容性设计测试作为默认要求
|
||||
|
||||
## Key Quotes
|
||||
> "Validates design decisions with real user data, not assumptions." — Agent Vibe
|
||||
> "Based on 25 user interviews and 300 survey responses, 80% of users struggled with..." — Communication Style
|
||||
|
||||
## Key Concepts
|
||||
- [[User Research]]:用户研究,通过定性定量方法理解用户行为
|
||||
- [[Usability Testing]]:可用性测试,验证设计决策
|
||||
- [[User Persona]]:用户画像,基于实证数据创建
|
||||
- [[User Journey Mapping]]:用户旅程映射,识别痛点
|
||||
- [[A/B Testing]]:A/B 测试,数据驱动决策
|
||||
- [[Behavioral Analytics]]:行为分析,识别行为模式
|
||||
|
||||
## Key Entities
|
||||
- [[UX Researcher]]:The Agency 项目中的用户体验研究智能体
|
||||
- [[The Agency]]:开源 AI 智能体集合项目
|
||||
|
||||
## Connections
|
||||
- [[UX Researcher]] ← provides_research ← [[Design Team]]
|
||||
- [[UX Researcher]] ← validates ← [[Product Decisions]]
|
||||
- [[UX Researcher]] ← employs ← [[User Research]]
|
||||
- [[UX Researcher]] ← employs ← [[Usability Testing]]
|
||||
45
wiki/sources/design-visual-storyteller.md
Normal file
45
wiki/sources/design-visual-storyteller.md
Normal file
@@ -0,0 +1,45 @@
|
||||
---
|
||||
title: "Visual Storyteller"
|
||||
type: source
|
||||
tags: [agent, design, storytelling]
|
||||
date: 2025-11-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/design/design-visual-storyteller.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:AI 智能体定义——Visual Storyteller(视觉故事讲述者),专注于视觉叙事、多媒体内容和品牌故事设计
|
||||
- 问题域:如何创建引人入胜的视觉故事,将复杂信息转化为具有情感吸引力的视觉内容
|
||||
- 方法/机制:通过视觉叙事框架、故事弧创作、数据可视化、跨平台适配实现视觉 storytelling
|
||||
- 结论/价值:为品牌和产品提供专业的视觉叙事能力,提升用户参与度和品牌认知
|
||||
|
||||
## Key Claims
|
||||
- Visual Storyteller 需确保每个视觉故事具有清晰的叙事结构(开端、发展、结局)
|
||||
- 视觉内容需考虑文化敏感性,确保国际市场的适应性
|
||||
- 跨平台视觉策略需适配 Instagram、YouTube、TikTok、LinkedIn、Pinterest 等不同平台
|
||||
- 成功指标:视觉内容参与率提升 50%,故事完成率 80%,品牌认知提升 35%
|
||||
|
||||
## Key Quotes
|
||||
> "Every visual story must have clear narrative structure (beginning, middle, end)" — 视觉叙事标准
|
||||
> "Transforms complex information into visual narratives that move people" — 核心设计理念
|
||||
|
||||
## Key Concepts
|
||||
- [[视觉叙事]]:通过视觉元素讲述故事的能力
|
||||
- [[故事弧创作]]:Beginning (setup), middle (conflict), end (resolution) 三段式结构
|
||||
- [[数据可视化]]:将复杂数据转化为易理解的视觉形式
|
||||
- [[跨平台适配]]:根据不同平台特性调整视觉内容
|
||||
|
||||
## Key Entities
|
||||
- [[The Agency]]:Visual Storyteller 所属的开源 AI 智能体集合项目
|
||||
- [[UX Architect]]:The Agency 中的技术架构与 UX 专家智能体
|
||||
- [[UX Researcher]]:用户体验研究智能体
|
||||
- [[UI Designer]]:UI 设计智能体
|
||||
- [[Brand Guardian]]:品牌战略与身份保护专家智能体
|
||||
|
||||
## Connections
|
||||
- [[The Agency]] ← contains ← [[Visual Storyteller]]
|
||||
- [[Visual Storyteller]] ← collaborates_with ← [[UX Architect]]
|
||||
- [[Visual Storyteller]] ← collaborates_with ← [[UX Researcher]]
|
||||
- [[Visual Storyteller]] ← collaborates_with ← [[UI Designer]]
|
||||
- [[Visual Storyteller]] ← collaborates_with ← [[Brand Guardian]]
|
||||
46
wiki/sources/design-whimsy-injector.md
Normal file
46
wiki/sources/design-whimsy-injector.md
Normal file
@@ -0,0 +1,46 @@
|
||||
---
|
||||
title: "Whimsy Injector Agent Personality"
|
||||
type: source
|
||||
tags: []
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/design/design-whimsy-injector.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:品牌体验中的趣味性(Whimsy)注入专家
|
||||
- 问题域:如何在保持专业性的同时为品牌体验添加个性化和乐趣
|
||||
- 方法/机制:通过微交互设计、 Easter Egg 彩蛋、游戏化系统、趣味微文案实现品牌差异化
|
||||
- 结论/价值:策略性趣味元素可提升用户参与度 40%+,增强品牌记忆度和情感连接
|
||||
|
||||
## Key Claims
|
||||
- 每个趣味元素必须服务功能或情感目的
|
||||
- 趣味设计必须无障碍包容(支持屏幕阅读器、减少运动偏好)
|
||||
- 品牌个性应贯穿从专业场景到休闲场景的全光谱
|
||||
- 趣味元素需性能优化,不能影响页面加载速度
|
||||
|
||||
## Key Quotes
|
||||
> "Every playful element must serve a functional or emotional purpose" — 设计原则核心
|
||||
|
||||
> "Ensure whimsy is appropriate for brand context and target audience" — 场景适配要求
|
||||
|
||||
> "User engagement with playful elements shows high interaction rates (40%+ improvement)" — 成功指标
|
||||
|
||||
## Key Concepts
|
||||
- [[Whimsy Taxonomy]]: 微妙趣味、交互式趣味、发现式趣味、情境式趣味四类
|
||||
- [[Micro-Interaction]]: 微交互设计系统,按钮反馈、加载动画、庆祝效果
|
||||
- [[Easter Egg]]: 彩蛋系统,Konami 代码、点击序列触发隐藏功能
|
||||
- [[Gamification]]: 游戏化系统,成就解锁、进度庆祝、社区建设
|
||||
|
||||
## Key Entities
|
||||
- [[The Agency]]: 开源 AI 智能体集合项目,Whimsy Injector 是其中 design 部门的设计专家
|
||||
|
||||
## Connections
|
||||
- [[design-ux-architect]] ← extends ← [[design-whimsy-injector]](UX Architect 提供技术框架,Whimsy Injector 注入品牌个性)
|
||||
|
||||
## Contradictions
|
||||
- 与 [[design-ux-architect]] 潜在冲突:
|
||||
- 冲突点:趣味性与可用性的平衡
|
||||
- 当前观点:Whimsy Injector 强调趣味优先但保持包容性
|
||||
- UX Architect 观点:可能更强调功能性和技术实现
|
||||
@@ -0,0 +1,56 @@
|
||||
---
|
||||
id: learning-sessions-cloud-transformation-programme-20230808-183322-meeting-recording
|
||||
title: "Learning Sessions Cloud Transformation Programme-20230808 183322 Meeting Recording"
|
||||
type: source
|
||||
tags: [Cloud Learning, Terraform, ECS, IaC, CTP]
|
||||
sources: []
|
||||
last_updated: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Cloud & DevOps/Public-Cloud-Learning-Sessions/03_Terraform/learning-sessions-cloud-transformation-programme-20230808-183322-meeting-recordi.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:ECS 基础设施即代码部署(ECS Deployment using Infrastructure as Code)
|
||||
- 问题域:云转型项目中的容器化服务部署与自动化管理
|
||||
- 方法/机制:CTP 自研 ECS Terraform 模块、支持动态扩缩容、自动修复、金丝雀部署
|
||||
- 结论/价值:通过 Terraform 模块化实现 ECS 标准化部署,降低运维复杂度
|
||||
|
||||
## Key Claims
|
||||
- ECS 是 AWS 专有容器编排服务,与 AWS 服务深度集成
|
||||
- 动态扩缩容是由于不可预测负载模式的关键技术需求
|
||||
- CTP ECS 模块基于 Gruntwork 仓库构建,支持 Docker 容器作为逻辑单元
|
||||
- 模块支持 Listener 模式实现集中化 ECS 管理
|
||||
|
||||
## Key Quotes
|
||||
> "Businesses have to thrive in the middle of all these challenges and it is forged by code." — 行业面临的不可预测性和敏捷性需求推动了基础设施即代码的采用
|
||||
|
||||
> "We have implemented the listener approach because we have seen many of the products are you know they are downloading the quotes from the grant work and using locally." — 采用 Listener 模式解决本地化使用问题
|
||||
|
||||
## Key Concepts
|
||||
- [[ECS (Elastic Container Services)]]:AWS 专有容器编排服务
|
||||
- [[Infrastructure as Code (IaC)]]:基础设施即代码,通过声明式配置管理云资源
|
||||
- [[CTP (Cloud Transformation Programme)]]:云转型计划
|
||||
- [[Terraform]]:跨平台基础设施即代码工具
|
||||
|
||||
## Key Entities
|
||||
- [[JP]]:CTP 讲师,负责 ECS 业务和技术背景介绍
|
||||
- [[Raja M]]:CTP SRE,负责 ECS 模块开发
|
||||
- [[Gruntwork]]:基础设施代码库,ECS 模块基于其仓库构建
|
||||
|
||||
## Connections
|
||||
- [[ECS (Elastic Container Services)]] ← supports ← [[Infrastructure as Code (IaC)]]
|
||||
- [[Terraform]] ← manages ← [[ECS (Elastic Container Services)]]
|
||||
- [[CTP (Cloud Transformation Programme)]] ← uses ← [[Gruntwork]]
|
||||
|
||||
## Prerequisites
|
||||
- VPC(虚拟私有云)
|
||||
- ELB 安全组
|
||||
- EFS 卷挂载
|
||||
- YAML 或 JSON 配置
|
||||
|
||||
## Integrations
|
||||
- AWS CloudWatch
|
||||
- Splunk
|
||||
- Grafana
|
||||
- Prometheus
|
||||
@@ -0,0 +1,55 @@
|
||||
---
|
||||
title: "Learning Sessions ECS Deployment using IAC - 20230808"
|
||||
type: source
|
||||
tags: [AWS, ECS, IaC, Terraform]
|
||||
date: 2026-04-14
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Cloud & DevOps/Public-Cloud-Learning-Sessions/03_Terraform/learning-sessions-ecs-deployment-using-iac-20230808-183322-meeting-recording.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:ECS(Elastic Container Services)基础设施即代码部署
|
||||
- 问题域:企业云原生转型中的容器化部署挑战
|
||||
- 方法/机制:Terraform 模块化部署、Listener 集中管理、自动扩展/修复、金丝雀部署
|
||||
- 结论/价值:通过 IaC 实现 ECS 容器服务的一致性部署和可重复性
|
||||
|
||||
## Key Claims
|
||||
- ECS 是 AWS 专有技术,与 AWS 服务深度集成,相比 EKS 或原生 Kubernetes 有独特优势和挑战
|
||||
- 企业必须通过代码应对不可预测性和敏捷性需求
|
||||
- Listener 方法实现了多产品的集中化 ECS 管理
|
||||
- 模块化部署支持自动扩展、自动修复和金丝雀部署
|
||||
|
||||
## Key Quotes
|
||||
> "Businesses have to thrive in the middle of all these challenges and it is forged by code." — JP
|
||||
> "We have implemented the listener approach because we have seen many of the products are downloading the quotes from the grant work and using locally." — Raja M
|
||||
|
||||
## Key Concepts
|
||||
- [[Infrastructure as Code (IaC)]]:通过代码实现基础设施管理的实践
|
||||
- [[Terraform]]:IaC 工具,用于声明式定义云资源
|
||||
- [[ECS (Elastic Container Services)]]:AWS 容器编排服务
|
||||
- [[Auto Scaling]]:根据负载自动调整计算资源
|
||||
- [[Auto Healing]]:自动检测并恢复故障实例
|
||||
- [[Canary Deployment]]:金丝雀部署,逐步替换旧版本
|
||||
- [[Listener Approach]]:集中管理 ECS 的架构模式
|
||||
- [[VPC]]:虚拟私有云,提供网络隔离
|
||||
- [[ELB]]:弹性负载均衡,分发网络流量
|
||||
- [[Security Group]]:安全组,控制入站和出站流量
|
||||
- [[EFS]]:弹性文件系统,提供共享存储
|
||||
|
||||
## Key Entities
|
||||
- [[JP]]:演讲者,讲解 ECS 业务和技术背景
|
||||
- [[Raja M]]:演讲者,讲解 CTP 和 SRE 内部 ECS 模块
|
||||
- [[CTP (Cloud Transformation Program)]]:云转型计划
|
||||
- [[AWS]]:云服务提供商
|
||||
- [[SRE]]:站点可靠性工程团队
|
||||
- [[Public Cloud Learning Sessions]]:OpenText 每周二的学习会议系列
|
||||
|
||||
## Connections
|
||||
- [[CTP (Cloud Transformation Program)]] ← developed_by ← [[ECS Module]]
|
||||
- [[ECS (Elastic Container Services)]] ← manages ← [[Docker Containers]]
|
||||
- [[Listener Approach]] ← orchestrates ← [[ECS Services]]
|
||||
- [[Terraform]] ← provisions ← [[ECS Infrastructure]]
|
||||
|
||||
## Contradictions
|
||||
- 无冲突记录
|
||||
80
wiki/sources/paid-media-auditor.md
Normal file
80
wiki/sources/paid-media-auditor.md
Normal file
@@ -0,0 +1,80 @@
|
||||
---
|
||||
title: "Paid Media Auditor"
|
||||
type: source
|
||||
tags: [agent, the-agency, paid-media, audit]
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/paid-media/paid-media-auditor.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:付费媒体审计智能体,评估 Google Ads、Microsoft Ads 和 Meta 广告账户
|
||||
- 问题域:广告账户结构、追踪配置、出价策略、创意素材、受众定位、竞争定位
|
||||
- 方法/机制:200+ 检查点系统化评估, severity 评分,影响估算
|
||||
- 结论/价值:提供可执行审计报告,识别 15-30% 效率提升机会
|
||||
|
||||
## Key Claims
|
||||
- Paid Media Auditor 方法论评估广告账户如同审计师审查财务报表
|
||||
- 200+ 检查点评估覆盖账户结构、追踪测量、出价预算、关键词定向、创意、购物广告、竞争定位、着陆页
|
||||
- 每个发现都包含 severity 级别、业务影响和具体修复方案
|
||||
- 审计可识别 15-30% 效率提升机会
|
||||
|
||||
## Key Quotes
|
||||
> "Methodical, detail-obsessed paid media auditor who evaluates advertising accounts the way a forensic accountant examines financial statements — leaving no setting unchecked, no assumption untested, and no dollar unaccounted for."
|
||||
|
||||
## Key Concepts
|
||||
- [[账户架构]]:广告活动组织结构设计
|
||||
- [[自动化出价策略]]:PPC 广告平台的智能出价优化机制
|
||||
- [[绩效最大化]]:Google Ads 自动化全渠道投放类型
|
||||
- 审计检查点:200+ 系统化评估维度
|
||||
- 严重性评分:critical, high, medium, low 四级
|
||||
|
||||
## Key Entities
|
||||
- [[The Agency]]:开源 AI 智能体集合项目
|
||||
- Google Ads:广告投放平台
|
||||
- Microsoft Ads:广告投放平台
|
||||
- Meta Ads:广告投放平台
|
||||
- GA4:Google Analytics 4
|
||||
- GTM:Google Tag Manager
|
||||
|
||||
## Connections
|
||||
- [[paid-media-ppc-strategist]] ← extends ← [[paid-media-auditor]]
|
||||
- [[paid-media-programmatic-buyer]] ← extends ← [[paid-media-auditor]]
|
||||
- [[The Agency]] ← contains ← [[paid-media-auditor]]
|
||||
|
||||
## Contradictions
|
||||
- 无
|
||||
|
||||
## Specialized Skills
|
||||
- 200+ point audit checklist execution with severity scoring
|
||||
- Impact estimation methodology
|
||||
- Platform-specific deep dives (Google Ads scripts, Microsoft Advertising import gap analysis, Meta Pixel/CAPI verification)
|
||||
- Executive summary generation
|
||||
- Historical trend analysis
|
||||
- Change history forensics
|
||||
- Compliance auditing for regulated industries
|
||||
|
||||
## Tooling
|
||||
- WebFetch, WebSearch, Read, Write, Edit, Bash
|
||||
- Google Ads MCP tools / API integrations (when available)
|
||||
- Google Ads scripts for automated data extraction
|
||||
- Platform API for live data extraction
|
||||
|
||||
## Use Cases
|
||||
- Full account audit before taking over management
|
||||
- Quarterly health checks
|
||||
- Competitive audit to win new business
|
||||
- Post-performance-drop diagnostic
|
||||
- Pre-scaling readiness assessment
|
||||
- Tracking and measurement validation
|
||||
- Annual strategic review
|
||||
- Compliance review for regulated verticals
|
||||
|
||||
## Success Metrics
|
||||
- 200+ checkpoints evaluated per account
|
||||
- 100% findings include specific fix instructions
|
||||
- 15-30% efficiency improvement opportunities identified
|
||||
- 3-5 business days turnaround
|
||||
- 80%+ implementation rate within 30 days
|
||||
- Measurable improvement within 60 days
|
||||
55
wiki/sources/paid-media-creative-strategist.md
Normal file
55
wiki/sources/paid-media-creative-strategist.md
Normal file
@@ -0,0 +1,55 @@
|
||||
---
|
||||
title: "Paid Media Ad Creative Strategist"
|
||||
type: source
|
||||
tags: [The Agency, Paid Media, AI Agent, Ad Creative, RSA]
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/paid-media/paid-media-creative-strategist.md]]
|
||||
|
||||
## Summary
|
||||
- **核心主题**:付费媒体广告创意策略智能体,负责广告文案撰写、响应式搜索广告(RSA)优化、素材组合设计和创意测试框架
|
||||
- **问题域**:Google Ads、Meta、Microsoft 和程序化广告平台的创意策略与优化
|
||||
- **方法/机制**:15-headline 策略设计、素材组 composition、A/B 测试框架、创意疲劳监测
|
||||
- **结论/价值**:在自动化出价环境下,创意是唯一真正可控的优化杠杆,每个标题、图片、视频都是待验证的假设
|
||||
|
||||
## Key Claims
|
||||
- 创意是自动化出价环境中剩余的最大优化杠杆——当算法控制出价、预算和定向时,创意是唯一真正可控的
|
||||
- 每个可能的标题/描述组合都必须符合语法和逻辑
|
||||
- 创意疲劳监测应通过展示阈值和 CTR 趋势来识别
|
||||
- A/B 测试应在 2-4 周内达到统计显著性
|
||||
|
||||
## Key Quotes
|
||||
> "Performance-oriented creative strategist who writes ads that convert, not just ads that sound good." — Agent 定义核心定位
|
||||
> "Every headline, description, image, and video is a hypothesis to be tested." — 创意测试理念
|
||||
|
||||
## Key Concepts
|
||||
- [[Responsive Search Ads (RSA)]]:响应式搜索广告,Google Ads 的 15-headline + 4-description 广告格式
|
||||
- [[Ad Strength]]:广告质量评分,Google 评估 RSA 质量的指标
|
||||
- [[Asset Group]]:素材组,Performance Max 广告系列的素材组织单元
|
||||
- [[Creative Fatigue]]:创意疲劳,广告展示过多后效果下降的现象
|
||||
- [[A/B Testing Framework]]:A/B 测试框架,系统化创意测试方法论
|
||||
|
||||
## Key Entities
|
||||
- [[The Agency]]:开源 AI 智能体集合项目
|
||||
- [[John Williams]](@itallstartedwithaidea):Agent 作者
|
||||
|
||||
## Connections
|
||||
- [[The Agency]] ← contains ← [[Paid Media Ad Creative Strategist]]
|
||||
- [[Paid Media Programmatic Display Buyer]] ← complements ← [[Paid Media Ad Creative Strategist]]
|
||||
- [[Paid Media PPC Campaign Strategist]] ← collaborates ← [[Paid Media Ad Creative Strategist]]
|
||||
- [[Search Query Analyst]] ← informs ← [[Paid Media Ad Creative Strategist]]
|
||||
|
||||
## Contradictions
|
||||
- 无已知冲突
|
||||
|
||||
---
|
||||
|
||||
## Related Agents (The Agency Paid Media)
|
||||
|
||||
- [[Paid Media Programmatic Display Buyer]]
|
||||
- [[Paid Media PPC Campaign Strategist]]
|
||||
- [[Paid Media Paid Social Strategist]]
|
||||
- [[Paid Media Auditor]]
|
||||
- [[Search Query Analyst]]
|
||||
46
wiki/sources/paid-media-paid-social-strategist.md
Normal file
46
wiki/sources/paid-media-paid-social-strategist.md
Normal file
@@ -0,0 +1,46 @@
|
||||
---
|
||||
title: "Paid Media Paid Social Strategist"
|
||||
type: source
|
||||
tags: [agent, the-agency, paid-media, advertising]
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/paid-media/paid-media-paid-social-strategist.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:The Agency 项目中的付费社交媒体广告智能体
|
||||
- 问题域:跨平台付费社交广告策略设计(Meta、LinkedIn、TikTok、Pinterest、X、Snapchat)
|
||||
- 方法/机制:全漏斗广告架构、受众工程、创意策略、转化归因
|
||||
- 结论/价值:为营销人员提供跨平台付费社交广告的自动化策略设计与优化
|
||||
|
||||
## Key Claims
|
||||
- 付费社交广告本质上是打断用户而非回答问题,因此创意和受众必须赢得注意力
|
||||
- 每个平台都有独特的用户行为、算法机制和创意要求,不能简单复用相同素材
|
||||
- 跨平台预算分配应基于渠道证据,而非单纯经验判断
|
||||
|
||||
## Key Concepts
|
||||
- [[Full-Funnel Advertising]]:从认知→参与→再营销→留存的完整漏斗结构
|
||||
- [[Audience Engineering]]:基于 Pixel、自定义受众、CRM列表的精准定向策略
|
||||
- [[Custom Audience]]:基于网站 Pixel、CRM 列表或 engagement 构建的自定义受众
|
||||
- [[Lookalike Audience]]:基于种子受众扩展的相似受众
|
||||
- [[Conversions API]]:服务器端转化事件跟踪机制
|
||||
|
||||
## Key Entities
|
||||
- [[The Agency]]:开源 AI 智能体集合项目,本智能体所属框架
|
||||
- [[John Williams]]:本智能体作者,@itallstartedwithaidea
|
||||
|
||||
## Connections
|
||||
- [[Paid Media PPC Campaign Strategist]] ← related_to ← [[Paid Media Paid Social Strategist]]
|
||||
- [[Paid Media Programmatic & Display Buyer]] ← related_to ← [[Paid Media Paid Social Strategist]]
|
||||
- [[Paid Media Auditor]] ← related_to ← [[Paid Media Paid Social Strategist]]
|
||||
|
||||
## Contradictions
|
||||
- 无
|
||||
|
||||
## Specialized Skills(本智能体特有)
|
||||
- Meta Advantage+ Shopping 和 app campaign 优化
|
||||
- LinkedIn ABM 与 CRM segments 同步
|
||||
- TikTok 创意趋势识别与快速适配
|
||||
- 跨平台受众 suppression 防止频次超限
|
||||
- iOS 14+ 隐私影响缓解(SKAdNetwork、聚合事件测量)
|
||||
64
wiki/sources/paid-media-ppc-strategist.md
Normal file
64
wiki/sources/paid-media-ppc-strategist.md
Normal file
@@ -0,0 +1,64 @@
|
||||
---
|
||||
title: "Paid Media PPC Campaign Strategist Agent"
|
||||
type: source
|
||||
tags: [agent, the-agency, paid-media, ppc, google-ads, amazon-ads]
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/paid-media/paid-media-ppc-strategist.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:付费媒体 PPC 智能体,负责大规模搜索、购物和效果最大化广告系列架构设计
|
||||
- 问题域:跨 Google、Microsoft、Amazon 广告平台的企业级账户管理
|
||||
- 方法/机制:账户架构设计、自动化出价策略选择、预算分配框架、跨平台活动规划
|
||||
- 结论/价值:实现 10K 到 10M+ 月预算的可扩展投放,平衡 ROAS 与增量增长
|
||||
|
||||
## Key Claims
|
||||
- 账户架构即战略 — 智能体将活动、广告组、受众和信号的完整系统视为驱动业务成果的整体
|
||||
- 自动化出价策略选择基于转化量与数据成熟度的动态评估
|
||||
- 绩效最大化广告系列需要资产组设计与信号优化协同
|
||||
- 增量测试框架(地理分割、留存组、匹配市场)是验证付费搜索真实价值的核心方法
|
||||
|
||||
## Key Quotes
|
||||
> "Account structure as strategy — not just keywords and bids, but how the entire system of campaigns, ad groups, audiences, and signals work together to drive business outcomes." — 定义核心方法论
|
||||
|
||||
> "Always prefer live API data over manual exports or screenshots." — 数据获取原则
|
||||
|
||||
## Key Concepts
|
||||
- [[账户架构 (Account Architecture)]]:活动结构设计、广告组分类、标签系统、跨数百活动的命名规范
|
||||
- [[自动化出价策略]]:tCPA、tROAS、Max Conversions、Max Conversion Value、组合出价策略
|
||||
- [[预算管理]]:预算分配框架、节奏模型、边际回报分析、增量测试、季节性预算调整
|
||||
- [[关键词策略]]:匹配类型策略、否定关键词架构、变体管理、广泛匹配+智能出价部署
|
||||
- [[绩效最大化 (Performance Max)]]:Google 广告的自动化全渠道投放类型,需要资产组设计和信号优化
|
||||
- [[受众策略]]:第一方数据激活、Customer Match、相似细分、受众排除、观察模式 vs 定向模式
|
||||
- [[跨平台规划]]:Google/Microsoft/Amazon 预算分配、平台特定功能利用、统一衡量方法
|
||||
- [[增量测试]]:地理分割、留存组、匹配市场等验证付费搜索真实增量的方法
|
||||
|
||||
## Key Entities
|
||||
- [[Google Ads]]:主要 PPC 广告平台,支持搜索、购物、效果最大化等多种活动类型
|
||||
- [[Microsoft Advertising]]:Google Ads 的替代平台,Bing 搜索流量
|
||||
- [[Amazon Ads]]:电商平台广告,适合产品推广和品牌建设
|
||||
- [[The Agency]]:开源 AI 智能体集合项目,定义专业化 AI Agent 规范
|
||||
- John Williams (@itallstartedwithaidea):Agent 作者
|
||||
|
||||
## Connections
|
||||
- [[The Agency]] ← includes ← [[Paid Media PPC Campaign Strategist Agent]]
|
||||
- [[Google Ads]] ← used_by ← [[Paid Media PPC Campaign Strategist Agent]]
|
||||
- [[Microsoft Advertising]] ← used_by ← [[Paid Media PPC Campaign Strategist Agent]]
|
||||
- [[Amazon Ads]] ← used_by ← [[Paid Media PPC Campaign Strategist Agent]]
|
||||
- [[账户架构 (Account Architecture)]] ← core_skill_of ← [[Paid Media PPC Campaign Strategist Agent]]
|
||||
- [[自动化出价策略]] ← core_skill_of ← [[Paid Media PPC Campaign Strategist Agent]]
|
||||
- [[绩效最大化 (Performance Max)]] ← campaign_type_of ← [[Paid Media PPC Campaign Strategist Agent]]
|
||||
|
||||
## Contradictions
|
||||
- 无明显冲突
|
||||
|
||||
## Specialized Skills
|
||||
- 分层活动架构(品牌、非品牌、竞争对手、征服)及其隔离策略
|
||||
- 绩效最大化资产组设计与信号优化
|
||||
- 购物广告 feed 优化与补充 feed 策略
|
||||
- DMA 和地理定向策略
|
||||
- 转化动作层级设计(主要 vs 次要、微观 vs 宏观)
|
||||
- Google Ads API 和 Scripts 规模化自动化
|
||||
- MCC 级别账户组合策略
|
||||
62
wiki/sources/paid-media-programmatic-buyer.md
Normal file
62
wiki/sources/paid-media-programmatic-buyer.md
Normal file
@@ -0,0 +1,62 @@
|
||||
---
|
||||
title: "Paid Media Programmatic & Display Buyer"
|
||||
type: source
|
||||
tags: [agent, the-agency, paid-media, programmatic, display-advertising]
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/paid-media/paid-media-programmatic-buyer.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:程序化购买与展示广告智能体,专注于付费媒体的投放策略与执行
|
||||
- 问题域:展示广告规模投放、程序化购买、ABM展示营销、合作伙伴媒体采购
|
||||
- 方法/机制:DSP平台管理、Google展示网络、受众定向、频率封顶、品牌安全验证
|
||||
- 结论/价值:提供完整的付费媒体购买工作流,从投放策略到效果优化全链路支持
|
||||
|
||||
## Key Claims
|
||||
- 程序化购买必须从浪费识别开始,而非扩张
|
||||
- 展示广告成功需要从覆盖、频率、可见度、品牌提升角度思考,而非仅看最后点击 CPA
|
||||
- 每次展示应触达正确的人、正确的语境、正确的频率
|
||||
|
||||
## Key Quotes
|
||||
> "Every impression should reach the right person, in the right context, at the right frequency." — 核心投放理念
|
||||
|
||||
> "Waste identification comes before expansion." — 投放优化原则
|
||||
|
||||
## Key Concepts
|
||||
- [[程序化购买]]:通过DSP平台自动化采购展示广告库存
|
||||
- [[Google展示网络]]:Google的广告网络,包含数百万网站和应用
|
||||
- [[DSP]]:需求方平台,用于程序化购买广告库存
|
||||
- [[ABM]]:基于账户的营销,精准定向特定企业账户
|
||||
- [[频率封顶]]:控制同一用户看到广告的次数上限
|
||||
- [[可见度标准]]:衡量广告实际被用户看到的指标(MRC标准要求70%+)
|
||||
- [[无效流量]]:机器人或非人类流量,需控制在3%以下
|
||||
|
||||
## Key Entities
|
||||
- [[The Agency]]:开源AI智能体集合项目,提供144+个跨12个部门的专业化AI Agent
|
||||
- [[DV360]]:Google的Demand Side Platform,程序化购买平台
|
||||
- [[The Trade Desk]]:第三方DSP平台
|
||||
- [[Amazon DSP]]:亚马逊需求方平台
|
||||
- [[Demandbase]]:ABM展示广告平台
|
||||
- [[6Sense]]:ABM和意图数据平台
|
||||
|
||||
## Connections
|
||||
- [[The Agency]] ← contains ← [[Paid Media Programmatic & Display Buyer]]
|
||||
- [[程序化购买]] ← uses ← [[DSP]]
|
||||
- [[DSP]] ← includes ← [[DV360]], [[The Trade Desk]], [[Amazon DSP]]
|
||||
- [[ABM]] ← implements ← [[Demandbase]], [[6Sense]]
|
||||
|
||||
## Contradictions
|
||||
- 与搜索广告思维冲突:
|
||||
- 搜索广告:关键词触发、精确意图、即时转化
|
||||
- 展示广告:覆盖优先、品牌建设、长期转化窗口(90天归因)
|
||||
- 当前观点:展示广告应独立评估,不应与搜索广告混合归因
|
||||
|
||||
## Success Metrics
|
||||
- 可见度:70%+
|
||||
- 无效流量:<3%(一般)、<1%(复杂)
|
||||
- 频率管理:每月3-7次
|
||||
- CPM效率:垂直基准的85%-115%范围内
|
||||
- ABM目标账户触达:60%+
|
||||
- 合作伙伴媒体ROI:90天内正 pipeline 归因
|
||||
48
wiki/sources/paid-media-tracking-specialist.md
Normal file
48
wiki/sources/paid-media-tracking-specialist.md
Normal file
@@ -0,0 +1,48 @@
|
||||
---
|
||||
title: "Paid Media Tracking & Measurement Specialist"
|
||||
type: source
|
||||
tags: [The Agency, Paid Media, AI Agent, Tracking, Measurement]
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/paid-media/paid-media-tracking-specialist.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:付费媒体追踪与测量专家智能体,负责构建付费媒体优化的数据基础
|
||||
- 问题域:跨平台追踪架构设计、转化归因、数据准确性验证
|
||||
- 方法/机制:GTM容器架构、GA4事件设计、服务端标记、转化去重、隐私合规
|
||||
- 结论/价值:确保每个转化被正确计数,每一分广告支出可衡量,为出价算法提供准确数据
|
||||
|
||||
## Key Claims
|
||||
- 精准追踪是付费媒体优化的基石,错误的转化计数不仅浪费数据,还会误导出价算法
|
||||
- 追踪错误会悄然累积,5%的差异会导致出价算法优化方向错误
|
||||
- 糟糕的追踪不如不追踪——误计的转化会误导算法优化错误结果
|
||||
|
||||
## Key Quotes
|
||||
> "If it's not tracked correctly, it didn't happen." — 本智能体的核心理念
|
||||
|
||||
> "Always cross-reference platform-reported conversions against the actual API data." — 始终交叉验证平台报告的转化数据与实际API数据
|
||||
|
||||
## Key Concepts
|
||||
- [[Tag Management]]:GTM容器架构、工作区管理、触发器/变量设计、自定义HTML标签、同意模式实施
|
||||
- [[GA4 Implementation]]:事件分类设计、自定义维度/指标、增强型测量配置、电商数据层实现、跨域追踪
|
||||
- [[Conversion Tracking]]:Google Ads转化操作、增强型转化(网页和Lead)、离线转化API导入、转化价值规则
|
||||
- [[Server-Side Tagging]]:GTM服务端容器部署、第一方数据收集、Cookie管理、服务端富化
|
||||
- [[Attribution]]:数据驱动归因模型配置、跨渠道归因分析、增量测量设计、营销组合模型输入
|
||||
- [[Meta CAPI]]:Meta转化API服务端配置、事件去重(event_id匹配)、域名验证、聚合事件测量配置
|
||||
|
||||
## Key Entities
|
||||
- [[The Agency]]:开源AI智能体集合项目,本智能体所属框架
|
||||
- [[John Williams]]:The Agency项目贡献者,付费媒体智能体系列作者
|
||||
- [[Paid Media PPC Campaign Strategist]]:付费媒体PPC广告策略智能体,本智能体的协同对象
|
||||
|
||||
## Connections
|
||||
- [[The Agency]] ← contains ← [[Paid Media Tracking & Measurement Specialist]]
|
||||
- [[Paid Media Tracking & Measurement Specialist]] ← supports ← [[Paid Media PPC Campaign Strategist]]
|
||||
- [[GTM]] ← implements ← [[Tag Management]]
|
||||
- [[GA4]] ← provides ← [[GA4 Implementation]]
|
||||
- [[Meta CAPI]] ← enables ← [[Conversion Tracking]]
|
||||
|
||||
## Contradictions
|
||||
- 与 [[Paid Media Auditor]] 可能存在分析角度差异:追踪专家关注数据采集准确性,审计师关注账户效率优化
|
||||
103
wiki/sources/sales-coach.md
Normal file
103
wiki/sources/sales-coach.md
Normal file
@@ -0,0 +1,103 @@
|
||||
---
|
||||
title: "Sales Coach"
|
||||
type: source
|
||||
tags: [agent, the-agency, sales, coaching]
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/sales/sales-coach.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:Sales Coach Agent 是一个销售 coaching 专家智能体,通过提问引导销售代表提升技能,而不是直接告知答案
|
||||
- 问题域:销售代表技能发展、pipeline 管理、交易策略、预测准确性
|
||||
- 方法/机制:使用 Socratic 方法、Richardson Sales Performance 框架、MEDDPICC 资格认证框架,通过结构化 coaching 实现行为改变
|
||||
- 结论/价值:结构化 coaching 是销售领导者最高杠杆的活动,每投入一小时 coaching 带来的收入回报超过任何其他活动
|
||||
|
||||
## Key Claims
|
||||
- 销售代表开发需要区分技能差距(不知道怎么做)和意愿差距(知道但不去做),coaching 修复技能,管理修复意愿
|
||||
- 每次 coaching 互动的最低要求是产出一个具体的、可行为的、可操作的收获
|
||||
- 交易准备是每次重要会议前的必备流程:目标、买家需求、我们的诉求、三个最可能的反对意见及处理方案
|
||||
- 预测准确性的核心问题不是"感觉如何",而是"需要什么条件才能成交,能否展示这些条件已满足的证据"
|
||||
|
||||
## Key Quotes
|
||||
> "A lost deal with disciplined process is more valuable than a lucky win. Because process compounds and luck does not." — 核心理念
|
||||
|
||||
> "What would you do differently if you could replay that moment?" — Socratic 提问示例
|
||||
|
||||
> "Every hour spent coaching returns more revenue than any hour spent in a forecast call." — coaching 的投资回报
|
||||
|
||||
## Key Concepts
|
||||
- [[Sales Coaching]]:通过结构化提问和行为反馈提升销售代表技能的专业方法
|
||||
- [[Pipeline Review]]:pipeline 审查流程,从审问式转变为 coaching 对话
|
||||
- [[Call Coaching]]:电话 coaching,通过具体行为反馈提升通话质量
|
||||
- [[Deal Strategy]]:交易策略,包括 deal prep 和 blameless debrief
|
||||
- [[Forecast Accuracy]]:预测准确性,通过可验证证据而非乐观主义进行承诺
|
||||
- [[MEDDPICC]]:销售资格认证框架,用于诊断交易风险
|
||||
|
||||
## Key Entities
|
||||
- [[Sales Discovery Coach]]:Sales Coach Agent 的相关智能体,专注于销售发现方法论
|
||||
- [[Richardson Sales Performance]]:销售绩效框架,用于评估和管理销售代表能力
|
||||
|
||||
## Connections
|
||||
- [[Sales Discovery Coach]] ← shares_methodology ← [[Sales Coaching]]
|
||||
- [[Sales Coach]] ← uses_framework ← [[MEDDPICC]]
|
||||
- [[Sales Coach]] ← uses_framework ← [[Richardson Sales Performance]]
|
||||
|
||||
## Contradictions
|
||||
- 无显著冲突
|
||||
|
||||
---
|
||||
|
||||
## Technical Deliverables
|
||||
|
||||
### Rep Coaching Plan Template
|
||||
```markdown
|
||||
# Coaching Plan: [Rep Name]
|
||||
|
||||
## Current Performance
|
||||
- **Quota Attainment (YTD)**: [%]
|
||||
- **Win Rate**: [%]
|
||||
- **Average Deal Size**: [$]
|
||||
- **Sales Cycle Length**: [days]
|
||||
- **Pipeline Coverage**: [Ratio]
|
||||
|
||||
## Skill Assessment
|
||||
| Competency | Current Level | Target Level | Gap |
|
||||
|-----------|--------------|-------------|-----|
|
||||
| Discovery quality | [1-5] | [1-5] | [Notes] |
|
||||
| Qualification rigor | [1-5] | [1-5] | [Notes] |
|
||||
| Objection handling | [1-5] | [1-5] | [Notes] |
|
||||
| Executive presence | [1-5] | [1-5] | [Notes] |
|
||||
| Closing / next-step commitment | [1-5] | [1-5] | [Notes] |
|
||||
| Forecast accuracy | [1-5] | [1-5] | [Notes] |
|
||||
|
||||
## Focus Areas (Max 3)
|
||||
### Focus 1: [Skill]
|
||||
- **Current behavior**: [Specific observed behavior]
|
||||
- **Target behavior**: [Specific behavioral target]
|
||||
- **Coaching actions**: [How you will develop this]
|
||||
- **Milestone**: [Observable indicator]
|
||||
- **Target date**: [When behavior becomes habitual]
|
||||
```
|
||||
|
||||
### Pipeline Review Framework
|
||||
```markdown
|
||||
# Pipeline Review: [Rep Name] — [Date]
|
||||
|
||||
## Portfolio Health
|
||||
- **Total Pipeline**: [$] across [#] deals
|
||||
- **Weighted Pipeline**: [$]
|
||||
- **Pipeline-to-Quota Ratio**: [X:1] (target 3:1+)
|
||||
|
||||
## Deal Inspection (Top 5 by Value)
|
||||
| Deal | Value | Stage | Age | Key Question | Risk |
|
||||
|------|-------|-------|-----|-------------|------|
|
||||
| [Deal] | [$] | [Stage] | [Days] | "What do we not know?" | [Red/Yellow/Green] |
|
||||
```
|
||||
|
||||
## Success Metrics
|
||||
- Team quota attainment exceeds 90%
|
||||
- Average win rate improves by 5+ percentage points within two quarters
|
||||
- Forecast accuracy within 10% of actual at monthly commit level
|
||||
- New rep ramp time decreases by 20% through structured onboarding
|
||||
50
wiki/sources/sales-discovery-coach.md
Normal file
50
wiki/sources/sales-discovery-coach.md
Normal file
@@ -0,0 +1,50 @@
|
||||
---
|
||||
title: "Sales Discovery Coach"
|
||||
type: source
|
||||
tags: [sales, discovery, methodology, agent]
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/sales/sales-discovery-coach.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:销售发现(Discovery)方法论指导,教导销售团队如何通过提问、当前状态映射、差距量化和电话结构来发现真正的购买动机
|
||||
- 问题域:销售团队在发现阶段常见的问题——急于推销、提问浅尝辄止、无法触及买家真实痛点
|
||||
- 方法/机制:三大发现框架(SPIN Selling、Gap Selling、Sandler Pain Funnel)+ 精英发现电话结构 + AECR 异议处理框架
|
||||
- 结论/价值:发现阶段是交易成败的关键,而非演示、提案或谈判阶段。深度发现能揭示买家的真实购买动机和紧迫感
|
||||
|
||||
## Key Claims
|
||||
- 发现阶段是交易成败的关键,而非演示、提案或谈判阶段
|
||||
- 暗示性问题(Implication Questions)是交易达成的关键,因为它激活了买家的损失厌恶
|
||||
- 预算异议几乎从来不是关于预算,而是关于买家是否相信价值超过成本
|
||||
- 最好的销售员说话少,买家应该占用 60% 或更多的时间
|
||||
- 购买决定是情绪化的决定,需要理性论证
|
||||
|
||||
## Key Quotes
|
||||
> "A deal with shallow discovery is a deal built on sand." — 核心观点
|
||||
> "Buyers will work harder to avoid a loss than to capture a gain." — 损失厌恶原理
|
||||
> "The VP who tells you 'we need better reporting' has a deeper truth: 'I'm presenting to the board in Q3 and I don't trust my numbers.'" — 表面需求背后的真实动机
|
||||
|
||||
## Key Concepts
|
||||
- [[SPIN Selling]]:Neil Rackham 提出的四步提问法(Situation、Problem、Implication、Need-Payoff)
|
||||
- [[Gap Selling]]:当前状态与期望状态之间的差距,差距越大紧迫感越强
|
||||
- [[Sandler Pain Funnel]]:三层漏斗,从表面症状到业务影响到个人/情感 stakes
|
||||
- [[AECR Framework]]:异议处理四步法(Acknowledge、Empathize、Clarify、Reframe)
|
||||
- [[Upfront Contract]]:前置合同,开场时设定议程、获得时间同意、授权提出棘手问题
|
||||
- [[60/40 Rule]]:买家应占用 60% 时间,销售员不超过 40%
|
||||
|
||||
## Key Entities
|
||||
- [[Neil Rackham]]:SPIN Selling 方法论创始人
|
||||
- [[Keenan]]:Gap Selling 方法论创始人
|
||||
- [[Sandler]]:Sandler Pain Funnel 方法论创始人
|
||||
|
||||
## Connections
|
||||
- [[SPIN Selling]] ← 核心方法论 ← [[Sales Discovery Coach]]
|
||||
- [[Gap Selling]] ← 核心方法论 ← [[Sales Discovery Coach]]
|
||||
- [[Sandler Pain Funnel]] ← 核心方法论 ← [[Sales Discovery Coach]]
|
||||
- [[AECR Framework]] ← 异议处理 ← [[Sales Discovery Coach]]
|
||||
- [[The Agency]] ← 属于 ← [[Sales Discovery Coach]]
|
||||
|
||||
## Contradictions
|
||||
- 无明显冲突
|
||||
48
wiki/sources/sales-proposal-strategist.md
Normal file
48
wiki/sources/sales-proposal-strategist.md
Normal file
@@ -0,0 +1,48 @@
|
||||
---
|
||||
title: "Proposal Strategist Agent"
|
||||
type: source
|
||||
tags: [The Agency, Sales, AI Agent, Proposal]
|
||||
date: 2026-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/sales/sales-proposal-strategist.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:销售提案策略智能体,将 RFP 响应转化为引人入胜的制胜叙事
|
||||
- 问题域:销售提案写作、竞标策略、竞争定位
|
||||
- 方法/机制:制胜主题开发(Win Theme)、三幕式提案叙事、执行摘要撰写、竞争定位策略
|
||||
- 结论/价值:通过结构化叙事和差异化定位,将技术优势转化为客户感知价值
|
||||
|
||||
## Key Claims
|
||||
- 制胜主题必须出现在执行摘要、解决方案叙事、案例研究和定价说明中,孤立的主题是不可见的主题
|
||||
- 提案不是合规练习,而是说服文档——在商品化市场中,叙事是差异化因素
|
||||
- 执行摘要是最关键的部分,许多评估者尤其是高级利益相关者只阅读此部分
|
||||
- 不要直接批评竞争对手,要通过框架化优势创造有机对比
|
||||
|
||||
## Key Quotes
|
||||
> "Proposals are won on clarity and lost on generics." — 核心信念
|
||||
> "Never write a generic proposal. If the buyer's name, challenges, and context could be swapped for another client without changing the content, the proposal is already losing." — 提案写作铁律
|
||||
|
||||
## Key Concepts
|
||||
- [[Win Theme]]:制胜主题,连接解决方案与买家最紧迫需求的核心陈述
|
||||
- [[Three-Act Proposal Narrative]]:三幕式提案叙事,理解挑战→解决方案旅程→转型状态
|
||||
- [[Executive Summary]]:执行摘要,不是提案的总结,而是放在最前面的结束论证
|
||||
- [[Competitive Positioning]]:竞争定位,通过优势框架创造有机对比而非负面攻击
|
||||
- [[Proposal Strategy]]:提案策略,将 RFP 响应转化为说服性叙事的方法论
|
||||
|
||||
## Key Entities
|
||||
- [[The Agency]]:开源 AI 智能体集合项目,汇集各类专业化 AI Agent
|
||||
|
||||
## Connections
|
||||
- [[Proposal Strategist]] ← is_agent_of ← [[The Agency]]
|
||||
- [[Win Theme]] ← core_concept_of ← [[Proposal Strategy]]
|
||||
- [[Three-Act Proposal Narrative]] ← framework_of ← [[Proposal Strategy]]
|
||||
- [[Executive Summary]] ← critical_section_of ← [[Proposal Strategy]]
|
||||
- [[Competitive Positioning]] ← enables ← [[Win Theme]]
|
||||
|
||||
## Contradictions
|
||||
- 与传统提案写作方法的冲突:
|
||||
- 冲突点:传统方法强调合规性,新方法强调叙事说服力
|
||||
- 当前观点:提案是说服文档,不是合规检查表
|
||||
- 对方观点:完整回答所有 RFP 要求即算完成
|
||||
55
wiki/sources/search-query-analyst.md
Normal file
55
wiki/sources/search-query-analyst.md
Normal file
@@ -0,0 +1,55 @@
|
||||
---
|
||||
title: "Search Query Analyst"
|
||||
type: source
|
||||
tags: [The Agency, Paid Media, AI Agent]
|
||||
date: 2025-04-20
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/Agent/agency-agents/paid-media/paid-media-search-query-analyst.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:付费搜索查询分析智能体,专门挖掘搜索词报告、构建否定关键词分类体系、识别查询意图差距
|
||||
- 问题域:付费搜索账户的信号噪音比优化、浪费支出识别、高意图流量放大
|
||||
- 方法/机制:N-gram 分析、查询聚类、意图分类、否定关键词决策树、查询雕刻
|
||||
- 结论/价值:通过系统性查询分析,可消除 10-20% 的非转化支出,覆盖 <5% 的无效展示
|
||||
|
||||
## Key Claims
|
||||
- 搜索查询优化不是一次性任务,而是持续系统
|
||||
- 每花费在无关查询上的一美元都是从转化查询中窃取的
|
||||
- 搜索词报告分析应成为每次查询分析的第一步
|
||||
- 否定关键词冲突率应为零
|
||||
|
||||
## Key Quotes
|
||||
> "Expert search query analyst who lives in the data layer between what users actually type and what advertisers actually pay for."
|
||||
|
||||
> "Every dollar spent on an irrelevant query is a dollar stolen from a converting one."
|
||||
|
||||
## Key Concepts
|
||||
- [[Search Term Analysis]]:大规模搜索词报告挖掘、模式识别、N-gram 分析、按意图聚类
|
||||
- [[Negative Keyword Architecture]]:分层否定关键词列表(账户级、活动级、广告组级)、共享否定列表、冲突检测
|
||||
- [[Intent Classification]]:将查询映射到买家意图阶段(信息型、导航型、商业型、交易型)
|
||||
- [[Match Type Optimization]]:近似变体影响分析、广泛匹配查询扩展审计、短语匹配边界测试
|
||||
- [[Query Sculpting]]:通过否定关键词和匹配类型组合将查询导向正确的活动/广告组
|
||||
- [[Waste Identification]]:按支出加权的无关性评分、零转化查询标记、高 CPC 低价值查询隔离
|
||||
- [[N-gram Frequency Analysis]]:大规模识别重复出现的无关修饰词
|
||||
- [[Search Query Optimization System]]:多因素评分,评估查询-广告-落地页一致性
|
||||
|
||||
## Key Entities
|
||||
- [[The Agency]]:开源 AI 智能体集合项目,发布本智能体的来源项目
|
||||
- Paid Media PPC Campaign Strategist:付费媒体 PPC 智能体,负责广告系列架构设计
|
||||
- Paid Media Programmatic Buyer:付费媒体程序化购买智能体
|
||||
- Paid Media Auditor:付费媒体审计智能体,负责账户效率评估
|
||||
- John Williams (@itallstartedwithaidea):本智能体作者
|
||||
|
||||
## Connections
|
||||
- [[Paid Media PPC Campaign Strategist]] ← same_category ← [[Search Query Analyst]]
|
||||
- [[Paid Media Programmatic Buyer]] ← same_category ← [[Search Query Analyst]]
|
||||
- [[Paid Media Auditor]] ← same_category ← [[Search Query Analyst]]
|
||||
- [[The Agency]] ← published ← [[Search Query Analyst]]
|
||||
- [[Search Term Analysis]] ← performed_by ← [[Search Query Analyst]]
|
||||
- [[Negative Keyword Architecture]] ← created_by ← [[Search Query Analyst]]
|
||||
- [[Intent Classification]] ← performed_by ← [[Search Query Analyst]]
|
||||
|
||||
## Contradictions
|
||||
- 与广泛匹配策略可能存在冲突:广泛匹配追求覆盖更多查询,但增加了无关查询风险,需要通过否定关键词平衡
|
||||
45
wiki/sources/the-agency-contributing.md
Normal file
45
wiki/sources/the-agency-contributing.md
Normal file
@@ -0,0 +1,45 @@
|
||||
---
|
||||
title: "Contributing to The Agency"
|
||||
type: source
|
||||
tags: [contributing, open-source, ai-agents]
|
||||
date: 2026-04-20
|
||||
source_file: raw/Agent/agency-agents/CONTRIBUTING.md
|
||||
---
|
||||
|
||||
## Summary
|
||||
- 核心主题:AI 智能体集合项目 The Agency 的贡献指南
|
||||
- 问题域:开源 AI 智能体设计规范、Pull Request 流程、代码风格
|
||||
- 方法/机制:Agent 文件结构、Persona vs Operations 分离原则、外部服务依赖声明
|
||||
- 结论/价值:提供系统化的 AI 智能体贡献框架,降低协作门槛
|
||||
|
||||
## Key Claims
|
||||
- The Agency 欢迎贡献者通过创建新智能体、改进现有智能体、分享成功案例、报告问题四种方式参与贡献
|
||||
- Agent 文件遵循 Persona(身份、沟通风格、规则)与 Operations(使命、交付物、工作流、指标)语义分离结构
|
||||
- 优秀的智能体具备六大特征:鲜明性格、明确交付物、成功指标、经过验证的工作流、学习记忆、真实场景测试
|
||||
- Pull Request 最快路径是提交单个 markdown 文件(一个智能体定义)
|
||||
- Agent 应优先依赖具有免费层级的外部服务,且服务仅作为增强而非必要条件
|
||||
|
||||
## Key Quotes
|
||||
> "Great agents have: Narrow, deep specialization, Distinct personality and voice, Concrete code/template examples, Measurable success metrics" — Agent 设计原则
|
||||
|
||||
> "The test: is this agent for the user, or for the vendor?" — 外部服务依赖判断标准
|
||||
|
||||
> "What's the sweet spot? One markdown file — a new or improved agent." — PR 最佳实践
|
||||
|
||||
## Key Concepts
|
||||
- [[AgentDesignPrinciples]]:智能体设计六大原则(Strong Personality、Clear Deliverables、Success Metrics、Proven Workflows、Learning Memory、Real-world Testing)
|
||||
- [[AgentFileStructure]]:Persona 与 Operations 语义分离的文件结构
|
||||
- [[ExternalServicesDeclaration]]:外部服务声明规范,通过 frontmatter 的 services 字段声明依赖
|
||||
- [[PersonaOperationsSplit]]:将智能体区分为身份相关(Persona)与操作相关(Operations)的组织方式
|
||||
|
||||
## Key Entities
|
||||
- [[TheAgency]]:开源 AI 智能体集合项目,汇集各类专业化 AI Agent
|
||||
- [[MSitarzewski]]:The Agency 项目维护者
|
||||
|
||||
## Connections
|
||||
- [[TheAgency]] ← has_contribution_guide ← [[TheAgencyContributing]]
|
||||
- [[AgentDesignPrinciples]] ← defines ← [[TheAgencyContributing]]
|
||||
- [[AgentFileStructure]] ← documented_in ← [[TheAgencyContributing]]
|
||||
|
||||
## Contradictions
|
||||
- 无显著冲突
|
||||
Reference in New Issue
Block a user