feat(wiki): ingest nexus-spatial-discovery — 8-agent product discovery exercise
Ingest raw/Agent/agency-agents/examples/nexus-spatial-discovery.md Source: 8 The Agency agents parallel discovery exercise for Nexus Spatial Products: AI spatial command center (SpatialAIOps category) Key findings: 2D-first/WebXR strategy, debugging as killer use case Entities created: Nexus-Spatial, CrewAI Concepts created: SpatialAIOps, Command-Theater-Interface, Debugging-Visualization, Semantic-Zoom, Growth-Loop
This commit is contained in:
61
wiki/concepts/Command-Theater-Interface.md
Normal file
61
wiki/concepts/Command-Theater-Interface.md
Normal file
@@ -0,0 +1,61 @@
|
||||
---
|
||||
title: "Command Theater Interface"
|
||||
type: concept
|
||||
tags: [spatial-computing, xr, interface-design, ux]
|
||||
last_updated: 2026-03-05
|
||||
---
|
||||
|
||||
## Definition
|
||||
|
||||
命令剧院(Command Theater)——以用户为中心的空间界面布局范式,以弧形剧场形式围绕用户组织内容。源自 [[nexus-spatial-discovery]] 中 [[XR-Interface-Architect]] 的设计。
|
||||
|
||||
## Architecture
|
||||
|
||||
```
|
||||
OVERVIEW CANOPY
|
||||
(pipeline topology)
|
||||
~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
/ \
|
||||
/ FOCUS ARC (120 deg) \
|
||||
/ primary node graph work \
|
||||
/________________________________\
|
||||
| |
|
||||
LEFT | USER POSITION | RIGHT
|
||||
UTILITY | (origin 0,0,0) | UTILITY
|
||||
RAIL | | RAIL
|
||||
|__________________________________|
|
||||
\ /
|
||||
\ SHELF (below sightline) /
|
||||
\ agent status, quick tools/
|
||||
\_________________________ /
|
||||
```
|
||||
|
||||
## Four Zones
|
||||
|
||||
| Zone | Position | Content |
|
||||
|------|----------|---------|
|
||||
| Focus Arc | 120°, 1.2-2.0m | Primary node graph workspace |
|
||||
| Overview Canopy | above, 2.5-4.0m | Miniature pipeline topology + health heatmap |
|
||||
| Utility Rails | left/right flanks | Agent library, monitoring, logs |
|
||||
| Shelf | below sightline, 0.8-1.0m | Run/stop, undo/redo, quick tools |
|
||||
|
||||
## Three-Layer Depth System
|
||||
|
||||
| Layer | Depth | Content | Opacity |
|
||||
|-------|-------|---------|---------|
|
||||
| Foreground | 0.8-1.2m | Active panels, inspectors, modals | 100% |
|
||||
| Midground | 1.2-2.5m | Node graph, connections, workspace | 100% |
|
||||
| Background | 2.5-5.0m | Overview map, ambient status | 40-70% |
|
||||
|
||||
## Design Rationale
|
||||
|
||||
- 用户处于中心,信息按重要性和使用频率分层
|
||||
- z-axis(深度)对应执行顺序:input nodes 最远,output nodes 最近
|
||||
- 减少晕动症:所有交互发生在 0.8-2.5m 范围
|
||||
- 与 [[Semantic-Zoom]] 配合实现渐进式空间复杂度
|
||||
|
||||
## Related Concepts
|
||||
|
||||
- [[Semantic-Zoom]]:与 Command Theater 配合的导航范式
|
||||
- [[SpatialAIOps]]:Command Theater 的应用场景
|
||||
- [[Nexus-Spatial]]:Command Theater 的产品实现
|
||||
41
wiki/concepts/Debugging-Visualization.md
Normal file
41
wiki/concepts/Debugging-Visualization.md
Normal file
@@ -0,0 +1,41 @@
|
||||
---
|
||||
title: "Debugging Visualization"
|
||||
type: concept
|
||||
tags: [spatial-computing, debugging, ux, multi-agent]
|
||||
last_updated: 2026-03-05
|
||||
---
|
||||
|
||||
## Definition
|
||||
|
||||
调试可视化(Debugging Visualization)——通过空间叠加将运行时追踪信息直接叠加在工作流结构之上,提供 3D 可视化的调试体验。
|
||||
|
||||
由 [[nexus-spatial-discovery]] 中 [[UX-Researcher]] 识别为 **杀手级用例(Killer Use Case)**。
|
||||
|
||||
## Why It's the Killer Use Case
|
||||
|
||||
1. **Real quantified pain point**:用户花费 40% 时间通过日志检查调试 Agent 失败([[nexus-spatial-discovery]] 用户画像数据)
|
||||
2. **2D tools handle it poorly**:现有 2D 工具无法有效可视化复杂的 Agent 执行追踪
|
||||
3. **3D uniquely suited**:工作流结构(节点+连接)+ 运行时数据(时间序列/调用栈/token消耗)天然适合 3D 叠加
|
||||
|
||||
## UX Research Finding
|
||||
|
||||
> "Debugging Is the Killer Use Case. Spatial overlay of runtime traces on workflow structure solves a real, quantified pain point that no 2D tool handles well." — [[UX-Researcher]]
|
||||
|
||||
[[Product-Trend-Researcher]] 和 [[XR-Interface-Architect]] 均独立得出相同结论。
|
||||
|
||||
## Implementation in Nexus Spatial
|
||||
|
||||
- 7 种 Agent 状态可视化:Idle / Queued / Running / Streaming / Completed / Error / Paused
|
||||
- 状态通过边缘光晕、内饰动画、声音、粒子系统区分
|
||||
- LOD 节点系统:hover 显示最近 I/O,selected 显示完整追踪
|
||||
|
||||
## Key Design Insight
|
||||
|
||||
空间计算对**结构性**任务(放置、连接、重排节点)增加价值,但对**参数性**任务(文本输入、配置)制造摩擦。界面必须无缝混合空间和 2D 模式——2D 面板锚定在空间位置。
|
||||
|
||||
## Related Concepts
|
||||
|
||||
- [[Command-Theater-Interface]]:调试可视化的界面容器
|
||||
- [[Semantic-Zoom]]:调试视图的导航模式
|
||||
- [[SpatialAIOps]]:调试可视化是 SpatialAIOps 的核心差异化
|
||||
- [[Nexus-Spatial]]:调试可视化作为核心功能
|
||||
49
wiki/concepts/Growth-Loop.md
Normal file
49
wiki/concepts/Growth-Loop.md
Normal file
@@ -0,0 +1,49 @@
|
||||
---
|
||||
title: "Growth Loop"
|
||||
type: concept
|
||||
tags: [growth, product, go-to-market, viral]
|
||||
last_updated: 2026-03-05
|
||||
---
|
||||
|
||||
## Definition
|
||||
|
||||
增长飞轮(Growth Loop)——产品内生的自我强化增长机制,用户行为直接产生新用户的引入,而非依赖外部广告或营销渠道。
|
||||
|
||||
由 [[nexus-spatial-discovery]] 中 [[Growth-Hacker]] 设计 4 大增长飞轮。
|
||||
|
||||
## The 4 Nexus Spatial Growth Loops
|
||||
|
||||
### 1. "Wow Factor" Demo Loop
|
||||
|
||||
**机制**:空间演示天然可分享。一键"Share Spatial Preview"生成 WebXR 链接或视频。
|
||||
**目标 K-factor**:0.3-0.5
|
||||
**核心洞察**:空间计算的分享性(immersive demos are inherently shareable)
|
||||
|
||||
### 2. Template Marketplace
|
||||
|
||||
**机制**:高级用户发布管道模板,可通过搜索发现,引流新注册
|
||||
**核心洞察**:用户既是消费者也是创造者,内容自增长
|
||||
|
||||
### 3. Collaboration Seat Expansion
|
||||
|
||||
**机制**:一个工程师采用 → 分享给队友 → 团队扩展到付费计划(Slack/Figma 增长模式)
|
||||
**核心洞察**:协作工具的网络效应
|
||||
|
||||
### 4. Integration-Driven Discovery
|
||||
|
||||
**机制**:在 LangChain、n8n、OpenAI/Anthropic 合作伙伴目录中列出
|
||||
**核心洞察**:嵌入用户已有工作流,成为默认选择
|
||||
|
||||
## Growth Loop vs Traditional Acquisition
|
||||
|
||||
| | Growth Loop | Traditional Acquisition |
|
||||
|--|------------|----------------------|
|
||||
| 依赖 | 产品内生 | 外部渠道 |
|
||||
| 规模效应 | 越用越强 | 边际递减 |
|
||||
| 成本 | 递减 | 递增 |
|
||||
| 可持续性 | 高 | 低 |
|
||||
|
||||
## Related Concepts
|
||||
|
||||
- [[Nexus-Spatial]]:Growth Loop 的应用场景
|
||||
- [[SpatialAIOps]]:空间分享性是"Wow Factor"飞轮的基础
|
||||
40
wiki/concepts/Semantic-Zoom.md
Normal file
40
wiki/concepts/Semantic-Zoom.md
Normal file
@@ -0,0 +1,40 @@
|
||||
---
|
||||
title: "Semantic Zoom"
|
||||
type: concept
|
||||
tags: [spatial-computing, navigation, ux, xr]
|
||||
last_updated: 2026-03-05
|
||||
---
|
||||
|
||||
## Definition
|
||||
|
||||
语义缩放(Semantic Zoom)——4 级导航范式,随用户缩放层级揭示不同粒度的信息,类似 iPad/iOS 地图的缩放交互但应用于工作流导航。
|
||||
|
||||
由 [[nexus-spatial-discovery]] 中 [[XR-Interface-Architect]] 定义。
|
||||
|
||||
## Four Levels
|
||||
|
||||
| Level | What You See | Use Case |
|
||||
|-------|-------------|---------|
|
||||
| Fleet View | All workflows as abstract shapes, color-coded by status | Overview, portfolio management |
|
||||
| Workflow View | Node graph with labels and connections | Building, editing workflows |
|
||||
| Node View | Expanded configuration, recent I/O, status metrics | Inspection, configuration |
|
||||
| Trace View | Full execution trace with data inspection | Debugging, root cause analysis |
|
||||
|
||||
## Design Principles
|
||||
|
||||
1. **Glanceable Before Inspectable**:关键信息在 2 秒内通过颜色/大小/运动/位置感知
|
||||
2. **Progressive Spatial Complexity**:新用户从接近 2D 开始,空间能力随信心增长逐步揭示
|
||||
3. **Smooth transitions**:Overview → Focus 600ms,Focus → Detail 500ms,最大 600ms
|
||||
|
||||
## Relation to Command Theater
|
||||
|
||||
- Fleet View 对应 Overview Canopy(大范围,高空俯瞰)
|
||||
- Workflow/Node View 对应 Focus Arc(主要工作区)
|
||||
- Trace View 对应 Utility Rails(详细日志/追踪面板)
|
||||
|
||||
## Related Concepts
|
||||
|
||||
- [[Command-Theater-Interface]]:Semantic Zoom 的空间容器
|
||||
- [[Debugging-Visualization]]:Trace View 层级的主要应用
|
||||
- [[SpatialAIOps]]:Semantic Zoom 是 SpatialAIOps 导航的核心模式
|
||||
- [[Nexus-Spatial]]:Semantic Zoom 作为核心导航范式
|
||||
47
wiki/concepts/SpatialAIOps.md
Normal file
47
wiki/concepts/SpatialAIOps.md
Normal file
@@ -0,0 +1,47 @@
|
||||
---
|
||||
title: "SpatialAIOps"
|
||||
type: concept
|
||||
tags: [spatial-computing, ai-operations, product-strategy, spatial-ai]
|
||||
last_updated: 2026-03-05
|
||||
---
|
||||
|
||||
## Definition
|
||||
|
||||
Spatial AI Operations(空间AI运营)——将空间计算引入 AI Agent 编排和监控的全新产品品类。Nexus Spatial 创造性定义了此品类,而非在已有的 AI 可观测性仪表板赛道竞争。
|
||||
|
||||
核心价值主张:将监控从"阅读仪表板"转变为"身临其境你的基础设施"。
|
||||
|
||||
## Core Characteristics
|
||||
|
||||
- **Immersive monitoring**:通过空间界面实时监控 AI Agent 执行,而非 2D 图表/日志
|
||||
- **3D orchestration**:通过 3D 节点图构建和编排 Agent 工作流
|
||||
- **Spatial collaboration**:团队进入共享 3D 空间协作进行事故响应
|
||||
- **Debugging overlay**:运行时追踪叠加于工作流结构的 3D 可视化调试
|
||||
|
||||
## Market Context
|
||||
|
||||
由 [[nexus-spatial-discovery]] 中的 [[Product-Trend-Researcher]] 和 [[Brand-Guardian]] 共同验证:
|
||||
|
||||
- AI 编排工具市场:$13.5B(22.3% CAGR)
|
||||
- 空间计算市场:$170-220B
|
||||
- **gap**:现有产品要么是空间型但非 AI,要么是 AI 型但仅 2D
|
||||
|
||||
## Canonical Use Case
|
||||
|
||||
Nexus Spatial 的产品定位:AI Agent 沉浸式 3D 命令中心
|
||||
- 3D 节点图可视化 Agent 管道
|
||||
- 实时监控空间面板
|
||||
- 3D 空间拖放构建工作流
|
||||
- 共享空间协作
|
||||
|
||||
## Related Concepts
|
||||
|
||||
- [[Spatial-Computing]]:底层空间计算技术基础
|
||||
- [[WebXR]]:SpatialAIOps 的主要分发平台
|
||||
- [[Command-Theater-Interface]]:SpatialAIOps 的核心界面模式
|
||||
- [[Debugging-Visualization]]:SpatialAIOps 的杀手级用例
|
||||
- [[Semantic-Zoom]]:SpatialAIOps 的导航范式
|
||||
|
||||
## Contradictions
|
||||
|
||||
- 与传统 [[Multi-Agent-System-Reliability]] 的监控方法:后者依赖日志/追踪面板,前者通过空间叠加提供直观可穿戴的监控体验
|
||||
29
wiki/entities/CrewAI.md
Normal file
29
wiki/entities/CrewAI.md
Normal file
@@ -0,0 +1,29 @@
|
||||
---
|
||||
title: "CrewAI"
|
||||
type: entity
|
||||
tags: [ai-agents, multi-agent, orchestration, open-source]
|
||||
last_updated: 2026-04-26
|
||||
---
|
||||
|
||||
## Overview
|
||||
|
||||
开源多 Agent 编排框架,100,000+ 开发者使用,以快速执行著称。
|
||||
|
||||
## Key Characteristics
|
||||
|
||||
- **CLI-first**:命令行优先,最小化可视化工具
|
||||
- **Rapid execution**:以速度为核心竞争优势
|
||||
- **Role-based agents**:基于角色的 Agent 设计(CEO/Manager/Worker 等)
|
||||
- **Task delegation**:任务委派和结果聚合
|
||||
|
||||
## Competitive Position
|
||||
|
||||
在 [[nexus-spatial-discovery]] 的竞争分析中:
|
||||
- **Strength**:100K+ 开发者生态,快速执行
|
||||
- **UX Gap**:CLI-first,最小化可视化工具,无法直观理解复杂 Agent 关系
|
||||
- **Implication**:Nexus Spatial 的 3D 可视化填补了 CrewAI 的可视化空白
|
||||
|
||||
## Related
|
||||
|
||||
- [[LangChain]] / [[n8n]]:同类竞争产品
|
||||
- [[Nexus-Spatial]]:SpatialAIOps 产品定位,填补 CrewAI 可视化缺口
|
||||
61
wiki/entities/Nexus-Spatial.md
Normal file
61
wiki/entities/Nexus-Spatial.md
Normal file
@@ -0,0 +1,61 @@
|
||||
---
|
||||
title: "Nexus Spatial"
|
||||
type: entity
|
||||
tags: [product, spatial-computing, ai-agents, spatial-aiops, the-agency]
|
||||
last_updated: 2026-03-05
|
||||
---
|
||||
|
||||
## Overview
|
||||
|
||||
AI Agent 沉浸式 3D 命令中心——一个 VisionOS + WebXR 应用,为 AI Agent 编排提供空间化的监控、构建和协作环境。
|
||||
|
||||
由 8 个 [[The-Agency]] 专业 Agent 并行协作规划([[nexus-spatial-discovery]]),10 分钟 wall-clock time 产出完整产品规划。
|
||||
|
||||
## Core Capabilities
|
||||
|
||||
- **3D Node Graph Visualization**:将 Agent 工作流可视化为 3D 节点图,数据流朝向用户
|
||||
- **Real-time Spatial Monitoring**:实时监控面板以空间面板形式展示
|
||||
- **Drag-and-Drop 3D Workflow Building**:3D 空间内拖放构建工作流
|
||||
- **Shared Spatial Collaboration**:多人进入共享 3D 空间协作
|
||||
- **AI Support Agent in Workspace**:支持代理以内嵌节点形式存在于用户空间工作区
|
||||
|
||||
## Pricing
|
||||
|
||||
| Tier | Price | Target |
|
||||
|------|-------|--------|
|
||||
| Explorer | Free | 开发者、solo builders(3 agents,WebXR viewer)|
|
||||
| Pro | $99/user/month | 小团队(25 agents,协作)|
|
||||
| Team | $249/user/month | 中型 AI 团队(unlimited agents,分析)|
|
||||
| Enterprise | Custom ($2K-10K/mo) | 大型企业(SSO,RBAC,on-prem,SLA)|
|
||||
|
||||
## Strategy
|
||||
|
||||
- **Phase 1 (Months 1-6)**:2D web dashboard with Three.js 2.5D → 50 paying teams,$60K MRR
|
||||
- **Phase 2 (Months 6-12)**:WebXR spatial mode → 200 teams,$300K MRR
|
||||
- **Phase 3 (Months 12-18)**:Native VisionOS app(仅在 spatial 需求验证后)→ 500 teams,$1M+ MRR
|
||||
|
||||
## Technical Architecture
|
||||
|
||||
- Orchestration engine:Rust(亚毫秒调度,零 GC 停顿)
|
||||
- WebXR client:TypeScript + React Three Fiber
|
||||
- VisionOS client:Swift 6 + SwiftUI + RealityKit
|
||||
- Collaboration:Yjs (CRDT) + WebRTC
|
||||
- 8-service architecture:Auth / Workspace / Workflow / Orchestration / Collaboration / Streaming / Plugin / Billing
|
||||
|
||||
## Key Insights from Discovery
|
||||
|
||||
1. **2D-first, spatial-second**:所有 8 个 Agent 独立得出相同结论
|
||||
2. **Debugging is the killer use case**:空间叠加调试追踪是 3D 真正优于 2D 的场景
|
||||
3. **WebXR over VisionOS**:~100 万 Vision Pro 安装量无法支撑业务;Safari 2025 年末采纳 WebXR
|
||||
4. **"War room" collaboration**:多人共享 3D 空间进行事故响应是最强的空间价值主张
|
||||
|
||||
## Related Entities
|
||||
|
||||
- [[SpatialAIOps]]:Nexus Spatial 定义的新品类
|
||||
- [[Product-Trend-Researcher]]:验证市场机会
|
||||
- [[Brand-Guardian]]:定义品牌定位
|
||||
- [[Backend-Architect]]:设计技术架构
|
||||
- [[Growth-Hacker]]:规划 GTM 策略
|
||||
- [[UX-Researcher]]:识别用户痛点
|
||||
- [[XR-Interface-Architect]]:设计空间界面
|
||||
- [[LangChain]] / [[CrewAI]] / [[n8n]]:主要竞争产品
|
||||
@@ -4,6 +4,16 @@
|
||||
- [Overview](overview.md) — living synthesis
|
||||
|
||||
## Sources
|
||||
- [2026-04-25] [Multi-Agent Workflow: Startup MVP with Persistent Memory](sources/workflow-with-memory.md)
|
||||
- [2026-04-25] [Multi-Agent Workflow: Landing Page Sprint](sources/workflow-landing-page.md)
|
||||
- [2026-04-25] [Multi-Agent Workflow: Startup MVP](sources/workflow-startup-mvp.md)
|
||||
- [2026-04-25] [Examples - Agency Multi-Agent Collaboration Showcase](sources/readme.md)
|
||||
- [2026-04-25] [Workflow Example: Book Chapter Development](sources/workflow-book-chapter.md)
|
||||
- [2026-04-25] [Executive Summary Generator Agent Personality](sources/support-executive-summary-generator.md)
|
||||
- [2026-04-25] [Finance Tracker Agent Personality](sources/support-finance-tracker.md)
|
||||
- [2026-04-25] [Support Infrastructure Maintainer Agent Personality](sources/support-infrastructure-maintainer.md)
|
||||
- [2026-04-25] [Support Responder Agent Personality](sources/support-support-responder.md)
|
||||
- [2026-04-25] [Analytics Reporter Agent Personality](sources/support-analytics-reporter.md)
|
||||
- [2026-04-25] [Support Legal Compliance Checker Agent Personality](sources/support-legal-compliance-checker.md)
|
||||
- [2026-04-25] [Accessibility Auditor Agent Personality](sources/testing-accessibility-auditor.md)
|
||||
- [2026-04-25] [Tool Evaluator Agent Personality](sources/testing-tool-evaluator.md)
|
||||
@@ -13,19 +23,19 @@
|
||||
- [2026-04-25] [Testing Reality Checker](sources/testing-reality-checker.md)
|
||||
- [2026-04-25] [Workflow Optimizer Agent Personality](sources/testing-workflow-optimizer.md)
|
||||
- [2026-04-25] [API Tester Agent Personality](sources/testing-api-tester.md)
|
||||
- [2026-04-25] [OpenCode Integration](sources/readme.md)
|
||||
- [2026-04-25] [OpenCode Integration](sources/readme.md)
|
||||
- [2026-04-25] [OpenCode Integration](sources/readme.md)
|
||||
- [2026-04-25] [OpenCode Integration](sources/readme.md)
|
||||
- [2026-04-25] [OpenCode Integration](sources/readme.md)
|
||||
- [2026-04-25] [OpenCode Integration](sources/readme.md)
|
||||
- [2026-04-25] [OpenCode Integration](sources/readme.md)
|
||||
- [2026-04-25] [OpenCode Integration](sources/readme.md)
|
||||
- [2026-04-25] [Examples - Agency Multi-Agent Collaboration Showcase](sources/readme.md)
|
||||
- [2026-04-25] [Examples - Agency Multi-Agent Collaboration Showcase](sources/readme.md)
|
||||
- [2026-04-25] [Examples - Agency Multi-Agent Collaboration Showcase](sources/readme.md)
|
||||
- [2026-04-25] [Examples - Agency Multi-Agent Collaboration Showcase](sources/readme.md)
|
||||
- [2026-04-25] [Examples - Agency Multi-Agent Collaboration Showcase](sources/readme.md)
|
||||
- [2026-04-25] [Examples - Agency Multi-Agent Collaboration Showcase](sources/readme.md)
|
||||
- [2026-04-25] [Examples - Agency Multi-Agent Collaboration Showcase](sources/readme.md)
|
||||
- [2026-04-25] [Examples - Agency Multi-Agent Collaboration Showcase](sources/readme.md)
|
||||
- [2026-04-25] [Backend Architect with Memory](sources/backend-architect-with-memory.md)
|
||||
- [2026-04-25] [OpenCode Integration](sources/readme.md)
|
||||
- [2026-04-25] [OpenCode Integration](sources/readme.md)
|
||||
- [2026-04-25] [OpenCode Integration](sources/readme.md)
|
||||
- [2026-04-25] [OpenCode Integration](sources/readme.md)
|
||||
- [2026-04-25] [Examples - Agency Multi-Agent Collaboration Showcase](sources/readme.md)
|
||||
- [2026-04-25] [Examples - Agency Multi-Agent Collaboration Showcase](sources/readme.md)
|
||||
- [2026-04-25] [Examples - Agency Multi-Agent Collaboration Showcase](sources/readme.md)
|
||||
- [2026-04-25] [Examples - Agency Multi-Agent Collaboration Showcase](sources/readme.md)
|
||||
- [2026-04-25] [Historian Agent Personality](sources/academic-historian.md)
|
||||
- [2026-04-25] [Academic Geographer](sources/academic-geographer.md)
|
||||
- [2026-04-25] [Academic Narratologist](sources/academic-narratologist.md)
|
||||
@@ -408,17 +418,7 @@
|
||||
- [2026-04-21] [marketing-book-co-author](sources/marketing-book-co-author.md) — (expected: wiki/sources/marketing-book-co-author.md — source missing)
|
||||
- [2026-04-21] [marketing-zhihu-strategist](sources/marketing-zhihu-strategist.md) — (expected: wiki/sources/marketing-zhihu-strategist.md — source missing)
|
||||
- [2026-04-21] [marketing-douyin-strategist](sources/marketing-douyin-strategist.md) — (expected: wiki/sources/marketing-douyin-strategist.md — source missing)
|
||||
- [2026-04-21] [nexus-spatial-discovery](sources/nexus-spatial-discovery.md) — (expected: wiki/sources/nexus-spatial-discovery.md — source missing)
|
||||
- [2026-04-21] [workflow-with-memory](sources/workflow-with-memory.md) — (expected: wiki/sources/workflow-with-memory.md — source missing)
|
||||
- [2026-04-21] [workflow-landing-page](sources/workflow-landing-page.md) — (expected: wiki/sources/workflow-landing-page.md — source missing)
|
||||
- [2026-04-21] [workflow-startup-mvp](sources/workflow-startup-mvp.md) — (expected: wiki/sources/workflow-startup-mvp.md — source missing)
|
||||
- [2026-04-21] [OpenCode Integration](sources/readme.md)
|
||||
- [2026-04-21] [workflow-book-chapter](sources/workflow-book-chapter.md) — (expected: wiki/sources/workflow-book-chapter.md — source missing)
|
||||
- [2026-04-21] [support-executive-summary-generator](sources/support-executive-summary-generator.md) — (expected: wiki/sources/support-executive-summary-generator.md — source missing)
|
||||
- [2026-04-21] [support-finance-tracker](sources/support-finance-tracker.md) — (expected: wiki/sources/support-finance-tracker.md — source missing)
|
||||
- [2026-04-21] [support-infrastructure-maintainer](sources/support-infrastructure-maintainer.md) — (expected: wiki/sources/support-infrastructure-maintainer.md — source missing)
|
||||
- [2026-04-21] [support-support-responder](sources/support-support-responder.md) — (expected: wiki/sources/support-support-responder.md — source missing)
|
||||
- [2026-04-21] [support-analytics-reporter](sources/support-analytics-reporter.md)
|
||||
- [2026-04-26] [Nexus Spatial: Full Agency Discovery Exercise](sources/nexus-spatial-discovery.md) — 8个专业Agent并行协作,产出AI空间指挥中心产品完整规划(市场验证、技术架构、品牌策略、GTM、UX研究、项目执行方案);验证SpatialAIOps新品类;确定2D先行、WebXR分发、调试为杀手级用例。
|
||||
- [2026-04-20] [security](sources/security.md) — (expected: wiki/sources/security.md — source missing)
|
||||
- [2026-04-20] [llm-wiki](sources/llm-wiki.md) — (expected: wiki/sources/llm-wiki.md — source missing)
|
||||
- [2026-04-20] [baoyu-skills](sources/baoyu-skills.md) — (expected: wiki/sources/baoyu-skills.md — source missing)
|
||||
@@ -441,10 +441,10 @@
|
||||
- [legal-billing-time-tracking](sources/legal-billing-time-tracking.md) — (expected: wiki/sources/legal-billing-time-tracking.md — source missing)
|
||||
- [legal-client-intake](sources/legal-client-intake.md) — (expected: wiki/sources/legal-client-intake.md — source missing)
|
||||
- [hospitality-guest-services](sources/hospitality-guest-services.md) — (expected: wiki/sources/hospitality-guest-services.md — source missing)
|
||||
- [OpenCode Integration](sources/readme.md)
|
||||
- [Examples - Agency Multi-Agent Collaboration Showcase](sources/readme.md)
|
||||
- [marketing-seo-specialist](sources/marketing-seo-specialist.md) — (expected: wiki/sources/marketing-seo-specialist.md — source missing)
|
||||
- [marketing-agentic-search-optimizer](sources/marketing-agentic-search-optimizer.md) — (expected: wiki/sources/marketing-agentic-search-optimizer.md — source missing)
|
||||
- [OpenCode Integration](sources/readme.md)
|
||||
- [Examples - Agency Multi-Agent Collaboration Showcase](sources/readme.md)
|
||||
- [finance-bookkeeper-controller](sources/finance-bookkeeper-controller.md) — (expected: wiki/sources/finance-bookkeeper-controller.md — source missing)
|
||||
- [finance-fpa-analyst](sources/finance-fpa-analyst.md) — (expected: wiki/sources/finance-fpa-analyst.md — source missing)
|
||||
- [finance-investment-researcher](sources/finance-investment-researcher.md) — (expected: wiki/sources/finance-investment-researcher.md — source missing)
|
||||
@@ -533,6 +533,8 @@
|
||||
- [engineering-ai-data-remediation-engineer](sources/engineering-ai-data-remediation-engineer.md) — (expected: wiki/sources/engineering-ai-data-remediation-engineer.md — source missing)
|
||||
|
||||
## Entities
|
||||
- [Nexus-Spatial](entities/Nexus-Spatial.md) — AI Agent 沉浸式 3D 命令中心;SpatialAIOps 新品类定义者
|
||||
- [CrewAI](entities/CrewAI.md) — 开源多 Agent 编排框架;100K+ 开发者;CLI-first,可视化能力有限
|
||||
- [Acemoglu](entities/Acemoglu.md)
|
||||
- [ACI-318](entities/ACI-318.md)
|
||||
- [Acronis](entities/Acronis.md)
|
||||
@@ -574,6 +576,7 @@
|
||||
- [AWS-Step-Functions](entities/AWS-Step-Functions.md)
|
||||
- [Axton](entities/Axton.md)
|
||||
- [Azure](entities/Azure.md)
|
||||
- [Backend-Architect](entities/Backend-Architect.md)
|
||||
- [BackendArchitect](entities/BackendArchitect.md)
|
||||
- [baoyu](entities/baoyu.md)
|
||||
- [BehavioralNudgeEngine](entities/BehavioralNudgeEngine.md)
|
||||
@@ -606,6 +609,7 @@
|
||||
- [CodeCrafters](entities/CodeCrafters.md)
|
||||
- [CodeWeaver](entities/CodeWeaver.md)
|
||||
- [containerd](entities/containerd.md)
|
||||
- [Content-Creator](entities/Content-Creator.md)
|
||||
- [Coze](entities/Coze.md)
|
||||
- [Cursor](entities/Cursor.md)
|
||||
- [Curve-Finance](entities/Curve-Finance.md)
|
||||
@@ -633,6 +637,7 @@
|
||||
- [Eurocode](entities/Eurocode.md)
|
||||
- [fireworks-tech-graph](entities/fireworks-tech-graph.md)
|
||||
- [Flux](entities/Flux.md)
|
||||
- [Frontend-Developer](entities/Frontend-Developer.md)
|
||||
- [frp](entities/frp.md)
|
||||
- [Gamma-AI](entities/Gamma-AI.md)
|
||||
- [GDPR](entities/GDPR.md)
|
||||
@@ -647,6 +652,7 @@
|
||||
- [GoogleAds](entities/GoogleAds.md)
|
||||
- [GoogleCloud](entities/GoogleCloud.md)
|
||||
- [Grafana](entities/Grafana.md)
|
||||
- [Growth-Hacker](entities/Growth-Hacker.md)
|
||||
- [Gruntwork](entities/Gruntwork.md)
|
||||
- [HashiCorp](entities/HashiCorp.md)
|
||||
- [Heather-Norris](entities/Heather-Norris.md)
|
||||
@@ -727,6 +733,7 @@
|
||||
- [OpenCode](entities/OpenCode.md)
|
||||
- [OpenManus](entities/OpenManus.md)
|
||||
- [OpenNotebook](entities/OpenNotebook.md)
|
||||
- [OrchestratorAgent](entities/OrchestratorAgent.md)
|
||||
- [PageLM](entities/PageLM.md)
|
||||
- [Perplexica](entities/Perplexica.md)
|
||||
- [Phenops-Team](entities/Phenops-Team.md)
|
||||
@@ -743,7 +750,10 @@
|
||||
- [RackNerd](entities/RackNerd.md)
|
||||
- [RAIT-09](entities/RAIT-09.md)
|
||||
- [Raj-Vardhan-Singh](entities/Raj-Vardhan-Singh.md)
|
||||
- [Rapid-Prototyper](entities/Rapid-Prototyper.md)
|
||||
- [Reality-Checker](entities/Reality-Checker.md)
|
||||
- [Recapio](entities/Recapio.md)
|
||||
- [RetroBoard](entities/RetroBoard.md)
|
||||
- [RichardFeynman](entities/RichardFeynman.md)
|
||||
- [rsvg-convert](entities/rsvg-convert.md)
|
||||
- [rsync](entities/rsync.md)
|
||||
@@ -759,6 +769,7 @@
|
||||
- [SMACs](entities/SMACs.md)
|
||||
- [SONY](entities/SONY.md)
|
||||
- [SparkryAI](entities/SparkryAI.md)
|
||||
- [Sprint-Prioritizer](entities/Sprint-Prioritizer.md)
|
||||
- [SRE-Team](entities/SRE-Team.md)
|
||||
- [Stable-Diffusion](entities/Stable-Diffusion.md)
|
||||
- [stacer](entities/stacer.md)
|
||||
@@ -784,7 +795,9 @@
|
||||
- [TweetClaw](entities/TweetClaw.md)
|
||||
- [TypeScript-Language-Server](entities/TypeScript-Language-Server.md)
|
||||
- [Ubuntu-Server](entities/Ubuntu-Server.md)
|
||||
- [UI-Designer](entities/UI-Designer.md)
|
||||
- [Uptime-Kuma](entities/Uptime-Kuma.md)
|
||||
- [UX-Researcher](entities/UX-Researcher.md)
|
||||
- [Veeam](entities/Veeam.md)
|
||||
- [Vibe-Kanban](entities/Vibe-Kanban.md)
|
||||
- [VictoriaMetrics](entities/VictoriaMetrics.md)
|
||||
@@ -810,6 +823,11 @@
|
||||
- [苏东坡](entities/苏东坡.md)
|
||||
|
||||
## Concepts
|
||||
- [SpatialAIOps](concepts/SpatialAIOps.md) — 空间AI运营新品类;将空间计算引入 AI Agent 编排和监控
|
||||
- [Command Theater Interface](concepts/Command-Theater-Interface.md) — 命令剧院界面;以用户为中心的弧形空间布局范式
|
||||
- [Debugging Visualization](concepts/Debugging-Visualization.md) — 调试可视化;空间叠加运行时追踪于工作流结构之上;杀手级用例
|
||||
- [Semantic Zoom](concepts/Semantic-Zoom.md) — 语义缩放;4级导航范式,随缩放层级揭示不同粒度信息
|
||||
- [Growth Loop](concepts/Growth-Loop.md) — 增长飞轮;产品内生的自我强化增长机制
|
||||
- [14种UML图](concepts/14种UML图.md)
|
||||
- [7种视觉风格系统](concepts/7种视觉风格系统.md)
|
||||
- [ABTesting](concepts/ABTesting.md)
|
||||
@@ -950,6 +968,7 @@
|
||||
- [Content-Creator](concepts/Content-Creator.md)
|
||||
- [Content-Hashing](concepts/Content-Hashing.md)
|
||||
- [Content-Ingestion](concepts/Content-Ingestion.md)
|
||||
- [Context-Passing](concepts/Context-Passing.md)
|
||||
- [Context-Substrate](concepts/Context-Substrate.md)
|
||||
- [Context-Window](concepts/Context-Window.md)
|
||||
- [Continuous-Delivery](concepts/Continuous-Delivery.md)
|
||||
@@ -1043,6 +1062,7 @@
|
||||
- [Feature-Flag](concepts/Feature-Flag.md)
|
||||
- [FeatureList](concepts/FeatureList.md)
|
||||
- [Federated-Access](concepts/Federated-Access.md)
|
||||
- [Feedback-Loop](concepts/Feedback-Loop.md)
|
||||
- [FIA-Framework](concepts/FIA-Framework.md)
|
||||
- [File-System-First-UI](concepts/File-System-First-UI.md)
|
||||
- [File-Watcher](concepts/File-Watcher.md)
|
||||
@@ -1156,13 +1176,15 @@
|
||||
- [Luhmann-四原则](concepts/Luhmann-四原则.md)
|
||||
- [Mackinder-Heartland-Theory](concepts/Mackinder-Heartland-Theory.md)
|
||||
- [Management-Pack](concepts/Management-Pack.md)
|
||||
- [Marketing-Attribution](concepts/Marketing-Attribution.md)
|
||||
- [MCPOnceAllAgents](concepts/MCPOnceAllAgents.md)
|
||||
- [MEDDPICC](concepts/MEDDPICC.md)
|
||||
- [Medical-Advertisement-Review](concepts/Medical-Advertisement-Review.md)
|
||||
- [Marketing-Attribution](concepts/Marketing-Attribution.md)
|
||||
- [MeetingNotes](concepts/MeetingNotes.md)
|
||||
- [Memory-Backend](concepts/Memory-Backend.md)
|
||||
- [Memory-Based-Handoff](concepts/Memory-Based-Handoff.md)
|
||||
- [MEMORY.md](concepts/MEMORY.md.md)
|
||||
- [Merge-Point](concepts/Merge-Point.md)
|
||||
- [MessageMatch](concepts/MessageMatch.md)
|
||||
- [Micro-Recovery](concepts/Micro-Recovery.md)
|
||||
- [Miping](concepts/Miping.md)
|
||||
@@ -1176,6 +1198,7 @@
|
||||
- [MTTD](concepts/MTTD.md)
|
||||
- [MTTR](concepts/MTTR.md)
|
||||
- [Multi-Account-Deployment](concepts/Multi-Account-Deployment.md)
|
||||
- [Multi-Agent-Orchestration](concepts/Multi-Agent-Orchestration.md)
|
||||
- [Multi-AgentHub](concepts/Multi-AgentHub.md)
|
||||
- [Multi-AI-Review](concepts/Multi-AI-Review.md)
|
||||
- [Multi-Channel-Delivery](concepts/Multi-Channel-Delivery.md)
|
||||
@@ -1207,6 +1230,8 @@
|
||||
- [Pain-Point-Mining](concepts/Pain-Point-Mining.md)
|
||||
- [Paper-Abstract-Batch-Fetching](concepts/Paper-Abstract-Batch-Fetching.md)
|
||||
- [Parallel-Agent-Execution](concepts/Parallel-Agent-Execution.md)
|
||||
- [Parallel-Agent-Work](concepts/Parallel-Agent-Work.md)
|
||||
- [Parallel-Kickoff](concepts/Parallel-Kickoff.md)
|
||||
- [Partial-Dependence-Plots](concepts/Partial-Dependence-Plots.md)
|
||||
- [Partition-Updates](concepts/Partition-Updates.md)
|
||||
- [Passive-Learning](concepts/Passive-Learning.md)
|
||||
@@ -1252,6 +1277,7 @@
|
||||
- [PUID-PGID](concepts/PUID-PGID.md)
|
||||
- [Pull-Request-Governance](concepts/Pull-Request-Governance.md)
|
||||
- [Purpose-Built-Databases](concepts/Purpose-Built-Databases.md)
|
||||
- [Quality-Gate](concepts/Quality-Gate.md)
|
||||
- [Quality-Scoring-Algorithm](concepts/Quality-Scoring-Algorithm.md)
|
||||
- [QualityAdjustedCoverage](concepts/QualityAdjustedCoverage.md)
|
||||
- [RAG](concepts/RAG.md)
|
||||
@@ -1277,9 +1303,9 @@
|
||||
- [ResponsiveSearchAds](concepts/ResponsiveSearchAds.md)
|
||||
- [Retrieval](concepts/Retrieval.md)
|
||||
- [Reviewer](concepts/Reviewer.md)
|
||||
- [RFM-Analysis](concepts/RFM-Analysis.md)
|
||||
- [Rightsizing](concepts/Rightsizing.md)
|
||||
- [Risk-Mitigation](concepts/Risk-Mitigation.md)
|
||||
- [RFM-Analysis](concepts/RFM-Analysis.md)
|
||||
- [ROI](concepts/ROI.md)
|
||||
- [Rollback-Rate](concepts/Rollback-Rate.md)
|
||||
- [Root-Cause-Analysis](concepts/Root-Cause-Analysis.md)
|
||||
@@ -1313,6 +1339,7 @@
|
||||
- [Semantic-Index-Infrastructure](concepts/Semantic-Index-Infrastructure.md)
|
||||
- [Semantic-Search](concepts/Semantic-Search.md)
|
||||
- [Semantic-Versioning](concepts/Semantic-Versioning.md)
|
||||
- [Sequential-Handoff](concepts/Sequential-Handoff.md)
|
||||
- [Sequential-Thinking](concepts/Sequential-Thinking.md)
|
||||
- [Serverless-Computing](concepts/Serverless-Computing.md)
|
||||
- [Service-Control-Policies-SCPs](concepts/Service-Control-Policies-SCPs.md)
|
||||
@@ -1376,6 +1403,7 @@
|
||||
- [Three-Tier-Review-Mechanism](concepts/Three-Tier-Review-Mechanism.md)
|
||||
- [ThreeActProposalNarrative](concepts/ThreeActProposalNarrative.md)
|
||||
- [TieredCampaignArchitecture](concepts/TieredCampaignArchitecture.md)
|
||||
- [Time-Boxing](concepts/Time-Boxing.md)
|
||||
- [Time-to-Market](concepts/Time-to-Market.md)
|
||||
- [Todoist-API](concepts/Todoist-API.md)
|
||||
- [Token-Light-Design](concepts/Token-Light-Design.md)
|
||||
|
||||
100
wiki/log.md
100
wiki/log.md
@@ -1,3 +1,38 @@
|
||||
## [2026-05-05] ingest | Nexus Spatial: Full Agency Discovery Exercise
|
||||
- Source file: Agent/agency-agents/examples/nexus-spatial-discovery.md
|
||||
- Status: ✅ 成功摄入
|
||||
- Summary: 8个 The Agency 专业 Agent 并行协作,完成 AI 空间指挥中心产品 Nexus Spatial 的完整规划(10分钟 wall-clock time)。涵盖:市场验证(AI 编排 $13.5B + 空间计算 $170-220B 交叉空白)、技术架构(8服务 Rust 编排引擎)、品牌策略(定义 [[SpatialAIOps]] 新品类)、GTM 计划(3阶段渐进:2D→WebXR→VisionOS)、UX 研究(调试为杀手级用例)、空间界面架构(命令剧院 + 7态节点系统)、项目执行(35周,5团队)。跨 Agent 独立共识:2D-first / WebXR 优先 / 调试 killer use case。
|
||||
- Concepts created: [[SpatialAIOps]](空间AI运营新品类)、[[Command-Theater-Interface]](命令剧院界面范式)、[[Debugging-Visualization]](调试可视化杀手级用例)、[[Semantic-Zoom]](4级语义缩放导航)、[[Growth-Loop]](增长飞轮)
|
||||
- Entities created: [[Nexus-Spatial]](AI Agent 沉浸式3D命令中心产品)、[[CrewAI]](竞争产品;CLI-first,可视化能力有限)
|
||||
- Source page: wiki/sources/nexus-spatial-discovery.md
|
||||
- Notes: index.md 原占位条目(2026-04-21 expected)已替换为完整摘要;overview.md 已追加 nexus-spatial-discovery 综合摘要条目;与 [[Multi-Agent-System-Reliability]] 在协调机制设计上的张力已记录于 Source Page Contradictions 节——Nexus Spatial 采用 Yjs CRDT 去中心化协作(面向人类协作者),Multi-Agent-System-Reliability 侧重自主 Agent 间显式协调(面向任务协调),场景不同但技术选择有潜在关联待研究。
|
||||
|
||||
## [2026-05-05] ingest | Examples - Agency Multi-Agent Collaboration Showcase
|
||||
- Source file: Agent/agency-agents/examples/README.md
|
||||
- Status: ✅ 成功摄入
|
||||
- Summary: agency-agents 示例目录——展示多个专业 Agent 如何协作完成真实任务的案例文档。核心价值:8 个专业 Agent 并行运行,产出连贯且相互引用的完整计划,无需协调开销;具体案例 Nexus-Spatial-Discovery 涵盖市场验证、技术架构、品牌策略、GTM 计划等 8 个领域。
|
||||
- Concepts created: 无(Multi-Agent-Collaboration 和 Nexus-Spatial-Discovery 均依赖实际 Source Page 内容,不在本文档范围内创建独立 Concept)
|
||||
- Entities created: 无(8 个参与 Agent 均已在 wiki 中有独立 Source Page;Nexus-Spatial-Discovery 已在 index.md 占位待实际案例文档摄入后完善)
|
||||
- Source page: wiki/sources/readme.md
|
||||
- Notes: index.md 原有 12 个过期 OpenCode Integration readme.md 占位条目,清理并替换为新条目;冲突检测:与 Multi-Agent-System-Reliability 在"是否需要显式协调机制"上的潜在冲突已记录于 Source Page Contradictions 节。
|
||||
|
||||
- Source file: Agent/agency-agents/support/support-infrastructure-maintainer.md
|
||||
- Status: ✅ 成功摄入
|
||||
- Summary: Infrastructure Maintainer——The Agency Support 部门的基础设施维护专家 Agent,核心理念"Keeps the lights on, the servers humming, and the alerts quiet",通过 Prometheus 监控/Terraform IaC/自动化灾备/安全合规实现运维自动化,确保 99.9%+ 服务可用性,降低 70%+ 人工运维任务,年度成本效率提升 20%+。与 Support Responder 和 Analytics Reporter 构成依赖链。
|
||||
- Concepts created: 无(Key Concepts 均为单来源特定概念/工具,不满足可独立复用阈值,均以 wikilink 形式记录于 Source Page Key Concepts 节)
|
||||
- Entities created: 无(Terraform/Prometheus/AWS-RDS/HashiCorp 等 Key Entities 均已在其他来源中存在,无需重复创建,均以 wikilink 形式记录于 Source Page Key Entities 节)
|
||||
- Source page: wiki/sources/support-infrastructure-maintainer.md
|
||||
- Notes: index.md 原占位条目(2026-04-21 source missing)已替换为完整摘要;overview.md Support 部门新增 support-infrastructure-maintainer 段落,位于 support-analytics-reporter 之后、Paid Media 部门之前;冲突检测:与 support-legal-compliance-checker 在"变更速度 vs 合规验证"上的张力已记录于 Source Page Contradictions 节,建议合规验证作为 CI/CD 流水线 Gate,不阻断常规变更但强制阻断高风险变更。
|
||||
|
||||
## [2026-05-05] ingest | Support Responder Agent Personality
|
||||
- Source file: Agent/agency-agents/support/support-support-responder.md
|
||||
- Status: ✅ 成功摄入
|
||||
- Summary: Support Responder——The Agency Support 部门的客户支持专员 Agent,核心理念"客户满意度优先于内部效率指标",通过多渠道(邮件/聊天/电话/社交媒体/应用内消息)提供全渠道支持,实现 85% 首次联系解决率和 2 小时首次响应 SLA。三级分层支持体系(Tier1 General/Tier2 Technical/Tier3 Specialists)通过明确升级标准和路由机制确保问题高效分流。配套 SupportAnalytics Python 框架(数据分析/趋势识别/改进建议/主动外展列表)和 KnowledgeBaseManager 知识库系统(文章创建/模板生成/交互式故障排除/内容优化)。成功指标:客户满意度 4.5+/5、首次联系解决率 80%+、SLA 合规率 95%+、知识库贡献减少相似工单 25%+。
|
||||
- Concepts created: 无(Omnichannel-Support/First-Contact-Resolution/Support-Tier-System 等 Key Concepts 均为单来源特定概念,不满足可独立复用阈值,均以 wikilink 形式记录于 Source Page Key Concepts 节)
|
||||
- Entities created: 无(Support-Responder-Agent/Support-Analytics-Dashboard/Knowledge-Base-Manager 等 Key Entities 均为单来源工具/交付物名称,不满足 ≥2 次出现阈值,均以 wikilink 形式记录于 Source Page Key Entities 节)
|
||||
- Source page: wiki/sources/support-support-responder.md
|
||||
- Notes: index.md 原占位条目(2026-04-21 source missing)已替换为完整摘要;overview.md Support 部门新增 support-support-responder 完整条目(含 Agent 间协作关系:Support Responder 产生工单数据→ Analytics Reporter 分析洞察;合规问题升级至 Legal Compliance Checker);冲突检测:与 support-legal-compliance-checker 在"问题解决优先 vs 合规优先"上的张力已记录于 Source Page Contradictions 节,建议分工:常规问题以 Support Responder 为主,涉及合规风险的问题升级至 Legal Compliance Checker。
|
||||
|
||||
## [2026-05-05] ingest | Analytics Reporter Agent Personality
|
||||
- Source file: Agent/agency-agents/support/support-analytics-reporter.md
|
||||
- Status: ✅ 成功摄入
|
||||
@@ -3488,3 +3523,68 @@
|
||||
- Entities created: GitHubCopilot
|
||||
- Source page: wiki/sources/github-copilot.md
|
||||
- Notes: 无内容冲突。index.md(Sources + Concepts + Entities)、overview.md(替换 Cursor Integration 段落为 GitHub Copilot Integration)、log.md 均已更新。Concept AgentFileFormat.md 已创建于 wiki/concepts/,Entity GitHubCopilot.md 已创建于 wiki/entities/。
|
||||
|
||||
## [2026-04-25] ingest | Finance Tracker Agent Personality
|
||||
- Source file: Agent/agency-agents/support/support-finance-tracker.md
|
||||
- Status: ✅ 成功摄入
|
||||
- Summary: Finance Tracker——The Agency Support 部门的企业财务规划与分析专家 Agent,核心理念"Keeps the books clean, the cash flowing, and the forecasts honest",通过 SQL 预算差异分析、Python 现金流预测(含季节性因子)、NPV/IRR 投资评估框架驱动财务决策,覆盖数据验证→预算编制→绩效监控→战略规划→合规审计全链路。关键指标:预算准确率 95%+、现金流预测准确率 90%+、平均投资回报 25%+、合规率 100%。
|
||||
- Concepts created: 无(Key Concepts 均为单来源特定财务概念如 BudgetVarianceAnalysis/CashFlowForecasting/NPV_IRR_Analysis/PaybackPeriod/RiskAssessment/FinancialCompliance,不满足可独立复用阈值,均以 wikilink 形式记录于 Source Page Key Concepts 节)
|
||||
- Entities created: 无(AgentsOrchestrator 已存在于 index;其余 Key Entities 均为单来源特定名称,不满足 ≥2 次出现阈值,均以 wikilink 形式记录于 Source Page Key Entities 节)
|
||||
- Source page: wiki/sources/support-finance-tracker.md
|
||||
- Notes: index.md 原占位条目(2026-04-21 source missing)已替换为完整摘要;overview.md 新增第 16 条(企业财务 Agent 全链路能力),位于第 15 条(Agent 去电通知 vs 来电接收)之后;冲突检测:未发现与现有 Wiki 页面的内容冲突;Entities/Concepts 无需新建独立页面。
|
||||
|
||||
## [2026-04-25] ingest | Support Executive Summary Generator Agent Personality
|
||||
- Source file: Agent/agency-agents/support/support-executive-summary-generator.md
|
||||
- Status: ✅ 成功摄入
|
||||
- Summary: Executive Summary Generator——The Agency Support 部门的咨询级执行摘要生成 Agent,融合麦肯锡 SCQA、BCG 金字塔原理、贝恩行动导向三大顶级咨询框架,将复杂商业输入转化为 325-475 词的高管级执行摘要,确保 C-suite 3 分钟内做出决策。每个发现必须量化,建议按 Critical/High/Medium 排序含负责人+时间线+预期结果。
|
||||
- Concepts created: 无(SCQA/Pyramid Principle/Action-Oriented Recommendations 均仅在本来源中出现,不满足 ≥ 2 次创建阈值,均以 wikilink 形式记录于 Source Page Key Concepts 节)
|
||||
- Entities created: 无(McKinsey/BCG/Bain 均仅在本来源中出现,不满足 ≥ 2 次创建阈值,均以 wikilink 形式记录于 Source Page Key Entities 节)
|
||||
- Contradictions detected: 无冲突。与 [[report-distribution-agent]] 存在协同关系(执行摘要生成后可由 Report Distribution Agent 分发),已在 Source Page Connections 节记录
|
||||
- Source page: wiki/sources/support-executive-summary-generator.md
|
||||
- Notes: index.md 原占位条目(2026-04-21 expected)已替换为完整摘要;overview.md Support 部门新增 [[support-executive-summary-generator]] 独立段落,位于 [[support-infrastructure-maintainer]] 之后,完善了 Support 部门数据→洞察→决策→分发的完整链路描述;Entities/Concepts 无需新建独立页面
|
||||
|
||||
## [2026-05-05] ingest | Workflow Example: Book Chapter Development
|
||||
- Source file: Agent/agency-agents/examples/workflow-book-chapter.md
|
||||
- Status: ✅ 成功摄入
|
||||
- Summary: Book Chapter Development——The Agency 的单 Agent 工作流示例,用于将录音、碎片化笔记或战略要点等原始素材转化为结构化的第一人称章节草稿,包含明确的编辑注释和修订循环。核心 Agent 为 Book Co-Author,输出五部分结构:目标、章节草稿、编辑注释、反馈循环、下一步。质量标准:保持第一人称声音、论点依附来源材料、删除泛化激励语言、以明确修订问题结尾。
|
||||
- Concepts created: 无(BookCoAuthor/EditorialRevisionLoop 等 Key Concepts 均为单来源框架性概念,不满足 ≥ 2 次创建阈值,均以 wikilink 形式记录于 Source Page Key Concepts 节)
|
||||
- Entities created: 无(Book Co-Author Agent 仅为本工作流的核心执行角色,未在其他来源出现 ≥ 2 次,无需新建独立 Entity 页面)
|
||||
- Contradictions detected: 与泛化 ghostwriting 服务的定位差异已记录于 Source Page Contradictions 节
|
||||
- Source page: wiki/sources/workflow-book-chapter.md
|
||||
- Notes: index.md 原占位条目(2026-04-21 expected)已替换为完整摘要;overview.md 无需修订(该来源属于 The Agency 内部框架文档,不涉及与其他 Agent/工作流的实质冲突);Entities/Concepts 无需新建独立页面,均以 wikilink 形式记录于 Source Page 相应节
|
||||
|
||||
## [2026-05-05] ingest | Multi-Agent Workflow: Startup MVP
|
||||
- Source file: Agent/agency-agents/examples/workflow-startup-mvp.md
|
||||
- Status: ✅ 成功摄入
|
||||
- Summary: 多 Agent 协作从创意到 MVP 的 4 周 7 步工作流示例([[workflow-startup-mvp]])。展示了 RetroBoard(远程团队回顾工具)4 周 MVP 开发的完整流水线:7 种专业 Agent(Sprint Prioritizer、UX Researcher、Backend Architect、Frontend Developer、Rapid Prototyper、Growth Hacker、Reality Checker)按顺序与并行方式协作。核心 4 大模式:Sequential Handoff(顺序交接)、Parallel Agent Work(并行工作)、Quality Gate(质量门控)、Context Passing(上下文传递)。
|
||||
- Entities created: RetroBoard、OrchestratorAgent、Sprint-Prioritizer、UX-Researcher、Backend-Architect、Frontend-Developer、Rapid-Prototyper、Growth-Hacker、Reality-Checker(共 9 个)
|
||||
- Concepts created: Sequential-Handoff、Parallel-Agent-Work、Quality-Gate、Context-Passing、Multi-Agent-Orchestration(共 5 个)
|
||||
- Source page: wiki/sources/workflow-startup-mvp.md
|
||||
- Notes: 无内容冲突;index.md 原占位条目(2026-04-21 expected)已替换为完整摘要;overview.md 已追加第 17 条综合摘要;所有 Entities 和 Concepts 均系首次出现,新建独立页面
|
||||
|
||||
## [2026-05-05] ingest | Nexus Spatial: Full Agency Discovery Exercise
|
||||
- Source file: Agent/agency-agents/examples/nexus-spatial-discovery.md
|
||||
- Status: ✅ 成功摄入
|
||||
- Summary: 8个 The Agency 专业 Agent 并行协作,完成 AI 空间指挥中心产品 Nexus Spatial 的完整规划(10分钟 wall-clock time)。涵盖:市场验证(AI 编排 $13.5B + 空间计算 $170-220B 交叉空白)、技术架构(8服务 Rust 编排引擎)、品牌策略(定义 [[SpatialAIOps]] 新品类)、GTM 计划(3阶段渐进:2D→WebXR→VisionOS)、UX 研究(调试为杀手级用例)、空间界面架构(命令剧院 + 7态节点系统)、项目执行(35周,5团队)。跨 Agent 独立共识:2D-first / WebXR 优先 / 调试 killer use case。
|
||||
- Concepts created: [[SpatialAIOps]](空间AI运营新品类)、[[Command-Theater-Interface]](命令剧院界面范式)、[[Debugging-Visualization]](调试可视化杀手级用例)、[[Semantic-Zoom]](4级语义缩放导航)、[[Growth-Loop]](增长飞轮)
|
||||
- Entities created: [[Nexus-Spatial]](AI Agent 沉浸式3D命令中心产品)、[[CrewAI]](竞争产品;CLI-first,可视化能力有限)
|
||||
- Source page: wiki/sources/nexus-spatial-discovery.md
|
||||
- Notes: index.md 原占位条目(2026-04-21 expected)已替换为完整摘要;overview.md 已追加 nexus-spatial-discovery 综合摘要条目;与 [[Multi-Agent-System-Reliability]] 在协调机制设计上的张力已记录于 Source Page Contradictions 节——Nexus Spatial 采用 Yjs CRDT 去中心化协作(面向人类协作者),Multi-Agent-System-Reliability 侧重自主 Agent 间显式协调(面向任务协调),场景不同但技术选择有潜在关联待研究。
|
||||
|
||||
|
||||
- Source file: Agent/agency-agents/examples/workflow-landing-page.md
|
||||
- Status: ✅ 成功摄入
|
||||
- Summary: 展示多 Agent 协作在一天内完成高转化率 landing page 的完整工作流,核心价值为 4 个可复用的设计模式:Parallel-Kickoff(并行启动)、Merge-Point(合并点)、Feedback-Loop(反馈循环)、Time-Boxing(时间盒)。4 个 Agent 角色:Content Creator(文案)、UI Designer(设计)、Frontend Developer(构建)、Growth Hacker(转化优化)。
|
||||
- Concepts created: Parallel-Kickoff, Merge-Point, Feedback-Loop, Time-Boxing
|
||||
- Entities created: Content-Creator, UI-Designer
|
||||
- Source page: wiki/sources/workflow-landing-page.md
|
||||
- Notes: index.md 原占位条目(2026-04-21 source missing)已替换为完整摘要;overview.md 新增 workflow-landing-page 段落,位于 The Agency 贡献指南之后、GitHub Copilot Integration 之前;冲突检测:与 workflow-startup-mvp 在时间粒度上互补(天 vs 周),无实质性冲突。
|
||||
|
||||
## [2026-04-26] ingest | Multi-Agent Workflow: Startup MVP with Persistent Memory
|
||||
- Source file: Agent/agency-agents/examples/workflow-with-memory.md
|
||||
- Status: ✅ 成功摄入
|
||||
- Summary: workflow-startup-mvp 的增强版——通过 MCP Memory Server 将手动复制粘贴交接升级为自动召回。核心机制:remember 存储 Agent 交付物(带项目名 + 接收方标签)、recall 自动召回上下文(无需人工粘贴)、rollback 回滚到上一个检查点(替代手动撤销)。Before/After 对比:会话超时丢失 / 多 Agent 需重复编译上下文 / QA 失败需手动描述问题 / 跨多天项目需重建上下文 → 跨会话持久 / 按标签共享 / 自动回滚 / 每次 pick up 继续。
|
||||
- Concepts created: Memory-Based-Handoff(基于记忆的交接模式,可抽象复用,已建独立 Concept 页面)
|
||||
- Entities created: 无(所有参与 Agent 和 RetroBoard 均已在 wiki 中存在,均以 wikilink 形式记录于 Source Page Key Entities 节)
|
||||
- Source page: wiki/sources/workflow-with-memory.md
|
||||
- Notes: index.md 原占位条目(2026-04-21 source missing)已替换为完整摘要;overview.md 新增 workflow-with-memory 段落,位于 Backend Architect with Memory 之后、multi-channel-assistant 之前;冲突检测:与 workflow-startup-mvp 在"是否需要人工复制粘贴"上存在表面冲突(实质为增强关系),已记录于 Source Page Contradictions 节——Memory 模式是原始工作流的增强层,Memory Server 可用时自动召回,不可使用时沿用手工粘贴策略。
|
||||
|
||||
@@ -9,6 +9,10 @@ The wiki covers two major multi-agent frameworks: **The Agency** (agency-agents)
|
||||
|
||||
**The Agency 贡献指南**([[contributing_zh-cn]] + [[contributing]] 英文原版):The Agency 项目贡献者指南——核心贡献方式:①创建全新智能体(8大分类:engineering/design/marketing/product/project-management/testing/support/spatial-computing/specialized);②优化现有智能体;③分享成功案例;④反馈问题。智能体设计五原则:**鲜明性格**(拒绝通用人设)、**明确交付物**(真实代码/模板)、**可量化指标**、**经过验证的工作流**、**学习记忆机制**。PR 流程包含提交前检查(真实场景测试、遵循模板、补充示例)、社区评审与迭代优化。属 [[Multi-Agent-System-Reliability]] 的智能体设计规范层,为 [[Multi-Agent-Team]] 提供标准化的智能体创建框架。
|
||||
|
||||
**[[nexus-spatial-discovery]]**(Nexus Spatial Discovery Exercise):8个 The Agency 专业 Agent 并行协作完成 AI 空间指挥中心产品完整规划的实战案例——10分钟 wall-clock time 产出完整规划。参与 Agent:产品趋势研究员(市场验证 + Vision Pro 现实核查)、后端架构师(8服务 Rust 架构)、品牌守护者(定义 [[SpatialAIOps]] 新品类)、增长黑客(3阶段 GTM + 增长飞轮)、支持应答者(AI 内嵌空间的差异化支持设计)、UX 研究员(识别调试为杀手级用例)、XR 界面架构师(命令剧院 + 7态节点系统)、项目牧羊人(35周时间线 + 5团队分配)。跨 Agent 独立共识:2D先行(WebXR分发) > VisionOS、品牌 > 调试 > 战情室协作 > 渐进披露。核心张力:Growth Hacker($29-59)与 Trend Researcher($99-249)定价分歧待 A/B 测试。属 [[Multi-Agent-System-Reliability]] 的多 Agent 协作规划层实践,展示了并行 Agent 发现可产出连贯、相互引用的完整计划。与 [[Multi-Agent-Team]](单团队多 Agent 架构)和 [[Agents-Orchestrator]](流水线编排)同属多 Agent 协作模式的不同维度。
|
||||
|
||||
**[[workflow-landing-page]]**:多 Agent 一天冲刺工作流——展示 Landing Page 场景下 4 个核心设计模式:**[[Parallel-Kickoff]]**(Content Creator + UI Designer 上午并行启动)、**[[Merge-Point]]**(Frontend Developer 等待两者完成)、**[[Feedback-Loop]]**(Growth Hacker 审查后 Frontend Developer 修改)、**[[Time-Boxing]]**(每个阶段严格时间盒:09:00→16:30)。与 [[workflow-startup-mvp]] 互补——后者以周为单位的长周期迭代,本工作流是单日冲刺的具体化实现。与 [[design-ui-designer]] 和 [[design-brand-guardian]] 共享 UI Designer 角色。
|
||||
|
||||
**GitHub Copilot Integration**([[github-copilot]]):The Agency 与 GitHub Copilot 的开箱即用集成——无需转换,Agency 的 `.md` + YAML frontmatter 格式与 GitHub Copilot 原生兼容。通过 `./scripts/install.sh --tool copilot` 一键安装,或手动复制到 `~/.github/agents/` 或 `~/.copilot/agents/` 目录。用户可在任意 Copilot 会话中通过名称激活特定 agent,如 `"Activate Frontend Developer and help me build a React component."`。与 [[readme|Cursor Integration]] 互补——后者项目级别生效,Copilot Integration 用户级别全局生效,共同构成 [[The Agency]] 的多 IDE 集成生态。
|
||||
|
||||
**Windsurf Integration**([[windsurf-integration]]):The Agency Agent roster 与 Windsurf 编辑器的集成方案——通过 `.windsurfrules` 文件将全部 Agent roster 聚合为单一规则文件,install.sh 脚本从项目根目录安装,项目级生效。Windsurf 中在 prompt 里按名称引用 Agent 即可激活(如 "Use the Frontend Developer agent to build this component.")。与 [[Cursor Integration]](.mdc 规则)和 [[Aider Integration]](CONVENTIONS.md)同为项目级 IDE 集成,机制相似,共同构成 The Agency 的多 IDE 覆盖体系。[[integrations-readme]] 已覆盖所有 11 种集成工具的概览。
|
||||
@@ -21,7 +25,9 @@ The wiki covers two major multi-agent frameworks: **The Agency** (agency-agents)
|
||||
|
||||
**MCP Memory Integration**([[mcp-memory-integration]]):The Agency 的 MCP Memory 集成方案——通过在 Agent 提示词中加入标准化的 Memory Integration 段落,为任意 Agent 赋予跨会话持久记忆能力,无需修改 Agent 代码。MCP Memory Server 提供四个核心工具:`remember`(存储决策/交付物快照)、`recall`(跨会话检索)、`rollback`(失败时回滚到上一个检查点)、`search`(跨 Agent 搜索记忆)。**Rollback 是杀手级功能**——当 QA 检查失败或架构决策出错时,直接恢复到已知良好状态而非从头重建。标签一致性是关键:每个记忆使用 Agent 名称和项目名称作为标签,确保 recall 可靠。与 [[specialized-mcp-builder]](构建 MCP Server)和 [[ai-memory-tools-two-camps]](AI 记忆工具全景分类)同属 The Agency MCP 生态的核心组成部分。
|
||||
|
||||
**Backend Architect with Memory**([[backend-architect-with-memory]]):The Agency 中具备持久记忆能力的后端架构师 Agent——专门负责可扩展系统设计、数据库架构、API 开发与云基础设施。核心记忆机制:会话启动时检索 `backend-architect` + 项目名标签的历史记忆,防止重复讨论已做决策;架构决策(选型数据库/定义 API 契约/选择通信模式)以标签化快照持久化;交付物(Schema/API 规范/架构文档)完成后主动标记接收方(如 `frontend-developer` + `api-spec`)供下游 Agent 查找;QA 失败时检索最近良好检查点回滚。作为 [[agents-orchestrator]] 调度的具体执行 Agent,通过 MCP Memory 实现多 Agent 协作中的上下文连续性。
|
||||
**Backend Architect with Memory**([[backend-architect-with-memory]]):The Agency 中具备持久记忆能力的后端架构师 Agent——专门负责可扩展系统设计、数据库架构、API 开发与云基础设施。核心记忆机制:会话启动时检索 `backend-architect` + 项目名标签的历史记忆,防止重复讨论已做决策;架构决策以标签化快照持久化;交付物完成后主动标记接收方供下游 Agent 查找;QA 失败时检索最近良好检查点回滚。作为 [[agents-orchestrator]] 调度的具体执行 Agent,通过 MCP Memory 实现多 Agent 协作中的上下文连续性。
|
||||
|
||||
**[[workflow-with-memory]]**(Multi-Agent Workflow: Startup MVP with Persistent Memory):[[workflow-startup-mvp]] 的增强版——通过 MCP Memory Server 将手动复制粘贴交接升级为自动召回,实现"记忆服务器作为粘合剂"。核心机制:`remember` 存储 Agent 交付物(带项目名 + 接收方标签)、`recall` 自动召回上下文(无需人工粘贴)、`rollback` 回滚到上一个检查点(替代手动撤销)。Before/After 对比:手动交接(会话超时丢失 / 多 Agent 需重复编译上下文 / QA 失败需手动描述问题 / 跨多天项目需重建上下文)→ Memory 模式(跨会话持久 / 按标签共享 / 自动回滚 / 每次 pick up 继续)。核心标签策略:所有记忆用项目名标签(如 retroboard),交付物额外用接收 Agent 标签(如 frontend-developer),这是 recall 正常工作的前提。Rollback 是 QA 失败恢复的核心:回滚到检查点而非手动追踪变化。与 [[workflow-startup-mvp]] 的关系:两者不冲突,Memory 模式是原始工作流的增强层——Memory Server 可用时自动召回;不可用时沿用原始工作流的手动粘贴策略。
|
||||
|
||||
**[[multi-channel-assistant]]**:基于 [[OpenClaw]] 的多渠道个人助理方案——以 Telegram Topic 路由为统一入口,整合 Google Workspace(gog)、Slack、Todoist、Asana,实现"说一句话完成全套工作"。核心价值:消除应用切换疲劳,AI 主动推送定时提醒(如每周垃圾清理、公司周报)。
|
||||
|
||||
@@ -114,8 +120,14 @@ The Agency 的 Support 部门涵盖数据分析、基础设施维护、法律合
|
||||
|
||||
**[[support-legal-compliance-checker]]**(Legal Compliance Checker):法律合规与监管专家 Agent——The Agency Support 部门的法律合规专家,负责确保企业运营、数据处理和内容创作符合多司法管辖区的法律法规和行业标准。核心理念:**合规优先**,任何业务流程变更前必须验证监管要求,将所有合规决策记录在法律推理和监管引用中。核心能力:四阶段工作流(监管格局评估 → 风险评估与差距分析 → 政策制定与实施 → 培训与文化建设);GDPR/CCPA/HIPAA/PCI-DSS/SOX/FERPA 多框架合规配置;Python PrivacyPolicyGenerator(多辖区隐私政策生成)和 ContractReviewSystem(关键词扫描风险评估)。交付物模板:Regulatory Compliance Assessment Report(合规评分体系,目标 98%+)。成功指标:98%+ 监管合规达标率、零监管处罚、95%+ 员工培训完成率、4.5+/5 合规文化评分。与 [[ComplianceAuditor]] 互补——Compliance Auditor 提供审计执行,Legal Compliance Checker 提供政策框架和风险评估;与 [[automation-governance-architect]] 协同——治理框架为政策制定提供结构化方法论;与 [[testing-reality-checker]] 存在张力:合规 Agent 达到监管标准即为合规,Reality Checker 要求压倒性视觉证明才授予生产就绪。
|
||||
|
||||
**[[support-support-responder]]**(Support Responder):客户支持专员 Agent——The Agency Support 部门的核心客户服务专家 Agent,专注于多渠道客户支持、问题解决和满意度优化,将每次支持互动转化为品牌正向体验。核心理念:**客户满意度优先于内部效率指标**,同理心沟通结合技术准确解决方案,适当升级超出权限的问题。核心能力:全渠道支持框架(邮件/聊天/电话/社交媒体/应用内消息,五渠道独立 SLA 配置)、三级分层支持体系(Tier1 General / Tier2 Technical / Tier3 Specialists,含升级标准和路由机制)、SupportAnalytics 数据分析框架(指标计算/趋势识别/改进建议/主动外展列表)、KnowledgeBaseManager 知识库系统(文章创建/模板生成/交互式故障排除/内容优化)。核心方法:四步工作流(客户查询分析与路由 → 问题调查与解决 → 客户跟进与成功测量 → 知识共享与流程改进)。成功指标:客户满意度 4.5+/5、首次联系解决率 80%+、SLA 合规率 95%+、知识库贡献减少相似工单 25%+。与 [[support-analytics-reporter]] 协同——Support Responder 产生工单原始数据,Analytics Reporter 将其转化为业务洞察;与 [[support-legal-compliance-checker]] 存在张力——常规问题以 Support Responder 为主,涉及合规风险的问题(数据删除/账户限制)升级至 Legal Compliance Checker;与 [[report-distribution-agent]] 协同——执行摘要生成基于支持数据分析结果。
|
||||
|
||||
**[[support-analytics-reporter]]**(Analytics Reporter):数据分析与商业智能专家 Agent——The Agency Support 部门的数据分析专家,负责将原始数据转化为可操作的业务洞察。核心使命:**用数据讲故事,用统计说话**。核心方法:四步工作流(数据发现验证 → 分析框架开发 → 洞察生成可视化 → 业务影响测量);技术栈覆盖 SQL(复杂业务指标 CTE 查询)、Python(pandas/scikit-learn 客户分层)、统计分析(p-value 显著性检验、回归、预测);交付物模板:Executive Dashboard(关键业务指标 + KPI 追踪)、Customer Segmentation Analysis(RFM 客户价值分层)、Marketing Attribution Dashboard(多触点归因 + ROI 分析)。核心原则:**数据质量优先**(分析前必须验证准确性、完整性,结论必须满足 p < 0.05 置信标准)、**业务影响聚焦**(所有分析必须连接到可量化的业务成果和可执行建议)、**可重现性保证**(版本控制 + 文档化的可重现分析工作流)。关键成功指标:分析准确率 95%+、分析建议实施率 70%+、利益相关者满意度 4.5+/5。与 [[support-finance-tracker]](财务追踪)和 [[report-distribution-agent]](报告分发)构成支持部门的数据→洞察→传递完整链路;与 [[sales-pipeline-analyst]] 共享数据分析能力,但前者提供通用 BI 框架,后者聚焦销售漏斗专项分析。属 [[Multi-Agent-System-Reliability]] 的数据驱动决策层,为多 Agent 销售系统提供统一的数据分析和业务洞察能力。
|
||||
|
||||
**[[support-infrastructure-maintainer]]**(Infrastructure Maintainer):基础设施维护专家 Agent——The Agency Support 部门的基础设施专家,负责确保系统可靠性、性能优化和技术运维管理,核心理念:**"Keeps the lights on, the servers humming, and the alerts quiet"**。核心能力:①监控告警系统(Prometheus + Grafana,CPU/内存/磁盘/服务可用性实时告警,99.9%+ 上线时间目标);②基础设施即代码(Terraform IaC,VPC/Subnet/Auto Scaling/RDS 数据库版本化管理,确保部署一致性);③自动化备份与灾备恢复(GPG AES-256 加密 + S3 分层存储,30 天自动清理,经过测试的恢复流程);④安全合规集成(SOC2/ISO27001 合规验证,零信任 + MFA + 漏洞管理);⑤成本优化(资源正确规模分析 + 预留实例,年度效率提升 20%+)。四步工作流:基础设施评估规划 → 监控实施 → 性能优化 → 安全合规验证。成功指标:上线时间 99.9%+、MTTR < 4 小时、70%+ 运维任务自动化、安全合规 100% 达标。**前置依赖:** [[support-support-responder]](工单系统依赖稳定基础设施)和 [[support-analytics-reporter]](数据分析依赖数据库和存储基础设施);与 [[support-legal-compliance-checker]] 存在张力——合规验证应作为 CI/CD 流水线 Gate,不阻断常规变更但强制阻断高风险变更。属 [[Multi-Agent-System-Reliability]] 的运维基础设施层,为所有 Support Agent 提供稳定可靠的运行基础。
|
||||
|
||||
**[[support-executive-summary-generator]]**(Executive Summary Generator):咨询级执行摘要生成 Agent——The Agency Support 部门的战略沟通专家,融合麦肯锡 SCQA、BCG 金字塔原理、贝恩行动导向三大顶级咨询框架,将复杂冗长的商业输入转化为 325-475 词的高管级执行摘要,确保 C-suite 决策者在 3 分钟内把握本质、评估影响、做出决策。核心理念:**洞察优先于信息,行动优先于描述**——每个关键发现必须包含量化数据点(≥1 个),不允许超越提供数据的假设,明确标记数据缺口。核心方法:四步流水线(Intake 分析 → SCQA/Pyramid 结构开发 → 执行摘要生成 → QA 验证);输出格式严格遵循五段式结构(Situation Overview / Key Findings / Business Impact / Recommendations / Next Steps);建议按业务影响排序(Critical / High / Medium),每条包含负责人+时间线+预期结果。成功指标:摘要阅读时间 <3 分钟、100% 发现含量化数据、325-475 词合规率 100%。与 [[support-analytics-reporter]] 协同——后者提供原始数据洞察,前者将其转化为高管可执行决策;与 [[support-legal-compliance-checker]] 协同——合规 Checker 的风险评估报告经 Executive Summary Generator 转化为高管行动建议;与 [[report-distribution-agent]] 协同——生成的执行摘要通过 Report Distribution Agent 分发给相关利益相关者。属 [[Multi-Agent-System-Reliability]] 的战略沟通层,为 C-suite 提供可执行的决策支撑文档。
|
||||
|
||||
### The Agency — Paid Media 部门
|
||||
The Agency 的 Paid Media 部门专注于企业级付费媒体策略与运营,涵盖 Google Ads、Microsoft Advertising、Amazon Ads 三大核心平台。
|
||||
|
||||
@@ -809,3 +821,9 @@ Key concepts: [[Django ORM]], [[Django REST Framework]], [[Django Admin 定制]]
|
||||
14. **SuperCall 沙盒 Persona vs 通用语音 Agent**:[[event-guest-confirmation]] 中使用的 [[SuperCall]] 强调独立沙盒设计——AI persona 只持有预设的 persona name、goal、opening line,无法访问外部系统;[[phone-based-personal-assistant]] 侧重通用个人助手场景,需要访问更多上下文。**[[Sandboxed Persona]]** 适用于确认类单一任务(安全、无注入风险);通用语音 Agent 适用于需要跨系统协调的复杂助手场景。
|
||||
|
||||
15. **Agent 去电通知 vs Agent 来电接收**:[[phone-call-notifications]] 中 Agent 主动向用户拨打电话通知(Agent → User),通话由 Agent 触发,用户是被动接收方;[[phone-based-personal-assistant]] 中用户主动呼叫 Agent(User → Agent),Agent 接听并提供助理服务。两者方向相反但互补——前者用于紧急告警、定时简报、重要事件通知,后者用于随时咨询、查询、执行任务。共同构成完整语音双向通信能力。
|
||||
|
||||
16. **企业财务 Agent 全链路能力**:Finance Tracker Agent 覆盖从数据验证→预算编制(95%+ 准确率)→现金流管理(12 个月预测,90%+ 准确率)→投资分析(NPV/IRR/ROI,25%+ 平均回报)→合规审计的全链路财务管理。核心差异化在于:多级审批 + 职责分离 + 完整审计跟踪。[[support-finance-tracker]]
|
||||
|
||||
17. **多 Agent 协作工作流关键模式**:[[workflow-startup-mvp]] 展示了 4 周 MVP 开发中 7 种专业 Agent 的协作模式。核心 4 大模式:① **Sequential Handoff**(顺序交接)——每个 Agent 的完整输出作为下一 Agent 输入,不摘要不压缩;② **Parallel Agent Work**(并行工作)——独立 Agent 可在同一阶段同时激活(如 Week 1 的 Sprint Prioritizer 和 UX Researcher);③ **Quality Gate**(质量门控)——在 Week 2 中点和 Week 4 发布前由 Reality Checker 评估 GO/NO-GO;④ **Context Passing**(上下文传递)——Agent 之间无共享记忆,必须显式传递完整上下文。未来引入 Orchestrator Agent 可替代手动传递,实现全自动化流水线。[[workflow-startup-mvp]]
|
||||
|
||||
|
||||
|
||||
76
wiki/sources/nexus-spatial-discovery.md
Normal file
76
wiki/sources/nexus-spatial-discovery.md
Normal file
@@ -0,0 +1,76 @@
|
||||
---
|
||||
title: "Nexus Spatial: Full Agency Discovery Exercise"
|
||||
type: source
|
||||
tags: [multi-agent, spatial-computing, product-discovery, vision-pro, webxr, ai-agents, spatial-aiops]
|
||||
date: 2026-03-05
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[Agent/agency-agents/examples/nexus-spatial-discovery.md]]
|
||||
|
||||
## Summary(用中文描述)
|
||||
- 核心主题:8个专业AI Agent并行协作,完成一个AI Agent空间指挥中心产品的完整发现与规划——从市场验证、技术架构、品牌策略、GTM计划到UX研究和项目执行方案。
|
||||
- 问题域:AI Agent编排工具市场($13.5B)+ 空间计算市场($170-220B)的交叉空白;现有AI工具均为平面2D界面,缺乏沉浸式3D命令中心。
|
||||
- 方法/机制:多Agent并行发现 → 各Agent独立研究 → 交叉Agent综合对齐 → 关键张力识别与决议;8个Agent(产品趋势研究员、后端架构师、品牌守护者、增长黑客、支持应答者、UX研究员、XR界面架构师、项目牧羊人)同步工作,约10分钟完成全栈规划。
|
||||
- 结论/价值:验证了"2D先行、空间其次"的渐进策略;调试是杀手级用例;WebXR优于VisionOS作为分发渠道;定义了全新品类SpatialAIOps;产出了35周完整执行计划。
|
||||
|
||||
## Key Claims(用中文描述)
|
||||
- AI编排工具($13.5B,22.3% CAGR)+ 空间计算($170-220B)存在交叉空白——现有产品要么是空间型但非AI,要么是AI型但仅2D。
|
||||
- Vision Pro全球仅~100万台,且销量从发布高点下降95%;不应以VisionOS为主推平台。
|
||||
- Safari于2025年末采纳WebXR Device API,2026年WebXR采用率增长40%——WebXR是理想的分发解锁渠道。
|
||||
- 调试(Debugging)是空间计算在AI工作流中的杀手级用例:工作流结构叠加运行时追踪的3D可视化解决了2D工具无法处理的问题。
|
||||
- 8个专业Agent并行工作可产出连贯、相互引用的完整计划,消除人工协调开销。
|
||||
|
||||
## Key Quotes
|
||||
> "Products are either spatial-but-not-AI-focused, or AI-focused-but-flat-2D. No product sits at the intersection." — Product Trend Researcher,核心竞争格局定性
|
||||
> "Debugging Is the Killer Use Case. Spatial overlay of runtime traces on workflow structure solves a real, quantified pain point that no 2D tool handles well." — UX Researcher,关键用例定位
|
||||
> "Do NOT lead with VisionOS. Lead with web, add WebXR, native VisionOS last." — Product Trend Researcher,产品平台策略核心结论
|
||||
> "Every agent independently arrived at this conclusion [2D-first, spatial-second]. Build a great web dashboard first, then progressively add spatial capabilities." — Cross-Agent Synthesis,跨Agent共识
|
||||
|
||||
## Key Concepts
|
||||
- [[SpatialAIOps]]:空间AI运营——将空间计算引入AI Agent编排和监控的全新品类;Nexus Spatial创造性地定义了此品类而非在拥挤的AI可观测性仪表板赛道竞争。
|
||||
- [[Multi-Agent-Orchestration]]:多Agent编排——Nexus Spatial的核心能力,涉及DAG执行、节点图可视化、实时监控和Agent生命周期管理。
|
||||
- [[WebXR]]:浏览器原生空间计算标准;Safari 2025年末采纳WebXR Device API,Android XR支持OpenXR,是Nexus Spatial的主要分发平台。
|
||||
- [[Command-Theater-Interface]]:命令剧院界面——以用户为中心的弧形空间布局:120度焦点弧(前)、概览天幕(上)、左右工具轨(侧)、工具架(下),通过三层深度系统组织内容。
|
||||
- [[Debugging-Visualization]]:调试可视化——空间叠加运行时追踪于工作流结构之上;UX研究发现这是2D工具处理最差、3D价值最高的用例。
|
||||
- [[Semantic-Zoom]]:语义缩放——4级导航(舰队视图→工作流视图→节点视图→追踪视图),随缩放层级揭示不同粒度的信息。
|
||||
- [[CRDT-Collaboration]]:CRDT(无冲突复制数据类型)+ Yjs + WebRTC,实现多人实时3D图协作编辑。
|
||||
- [[Rust-Orchestration]]:Rust实现的编排引擎——亚毫秒调度、零GC停顿、内存安全;用于AI Agent沙箱和关键路径DAG执行。
|
||||
- [[Gaze-Voice-Hands]]:注视+语音+手势三模态输入——空间计算的基础交互范式,支持双手解放和免手操作。
|
||||
- [[Growth-Loop]]:增长飞轮——"Wow Factor"演示分享循环(K=0.3-0.5)、模板市场、协作席位扩展、集成驱动发现。
|
||||
|
||||
## Key Entities
|
||||
- [[Nexus-Spatial]]:产品名——AI Agent沉浸式3D命令中心,VisionOS + WebXR应用;核心功能:3D节点图可视化Agent管道、实时监控空间面板、3D空间拖放构建工作流、共享空间协作;定价:Explorer免费、Pro $99/月、Team $249/月、Enterprise定制;战略定位:SpatialAIOps新品类创造者。
|
||||
- [[Product-Trend-Researcher]]:产品趋势研究员——The Agency执行董事Agent;发现Vision Pro市场数据,重新定义平台策略(WebXR优先于VisionOS);验证Nexus Spatial进入条件性通过(2D先行)。
|
||||
- [[Backend-Architect]]:后端架构师——The Agency工程Agent;设计了8服务架构(客户端→API网关→服务层→数据层→AI提供商层),推荐Rust用于编排引擎。
|
||||
- [[Brand-Guardian]]:品牌守护者——The Agency设计Agent;定义了"SpatialAIOps"新品类定位;设计了6色品牌系统和品牌个性(权威/沉稳/先锋/精确/亲和)。
|
||||
- [[Growth-Hacker]]:增长黑客——The Agency增长Agent;设计了3阶段GTM策略(创始人主导→开发者社区→企业);设计了4大增长飞轮。
|
||||
- [[Support-Responder]]:支持应答者——The Agency支持Agent;设计了支持分级体系;创新性地将AI支持代理嵌入用户空间工作区作为核心差异化。
|
||||
- [[UX-Researcher]]:UX研究员——The Agency研究Agent;识别了4个核心用户画像;发现调试是杀手级用例;提出空间增加价值/制造摩擦的临界点分析。
|
||||
- [[XR-Interface-Architect]]:XR界面架构师——The Agency空间计算Agent;设计了命令剧院布局和7态Agent节点系统;规划了3 LOD节点渲染和舒适度措施。
|
||||
- [[Project-Shepherd]]:项目牧羊人——The Agency项目管理Agent;制定了35周时间线(2026年3月-11月)和5个专项团队;识别了5大风险和$121.5K-$155.5K非人员预算。
|
||||
- [[LangChain]]:AI编排框架竞争对手;$39/用户/月;弱点:平面仪表板,大规模图不可读。
|
||||
- [[CrewAI]]:AI编排框架竞争对手;100K+开发者;弱点:CLI优先,可视化工具有限。
|
||||
- [[n8n]]:可视化工作流构建器;$20-50/月;弱点:2D画布无法处理Agent关系。
|
||||
- [[Apple-Vision-Pro]]:空间计算平台;全球~100万台(销量从高点下降95%);Apple战略转向轻量AR眼镜;~3,000个VisionOS原生应用;结论:不应作为主推平台。
|
||||
|
||||
## Connections
|
||||
- [[SpatialAIOps]] ← defines ← [[Nexus-Spatial]]
|
||||
- [[Product-Trend-Researcher]] → validated → [[Nexus-Spatial]] (CONDITIONAL GO)
|
||||
- [[UX-Researcher]] → identified → [[Debugging-Visualization]] (killer use case)
|
||||
- [[Backend-Architect]] → designed → [[Rust-Orchestration]]
|
||||
- [[XR-Interface-Architect]] → designed → [[Command-Theater-Interface]]
|
||||
- [[Growth-Hacker]] → planned → [[Growth-Loop]] strategy
|
||||
- [[Nexus-Spatial]] ← depends_on ← [[Rust-Orchestration]]
|
||||
- [[Nexus-Spatial]] ← depends_on ← [[WebXR]] (primary distribution)
|
||||
- [[Nexus-Spatial]] ← depends_on ← [[CRDT-Collaboration]]
|
||||
- [[Nexus-Spatial]] ← depends_on ← [[Multi-Agent-Orchestration]]
|
||||
- [[Nexus-Spatial]] ← depends_on ← [[Semantic-Zoom]] navigation
|
||||
- [[Nexus-Spatial]] ← depends_on ← [[Gaze-Voice-Hands]] interaction
|
||||
|
||||
## Contradictions
|
||||
- 与 [[Multi-Agent-System-Reliability]] 在协调机制设计上的张力:
|
||||
- 冲突点:Multi-Agent-System-Reliability强调需要显式协调机制(共享状态、仲裁者模式)防止多Agent冲突;Nexus-Spatial-Discovery采用Yjs CRDT实现无冲突实时协作,无需中央协调者。
|
||||
- 当前观点:Nexus-Spatial选择去中心化CRDT方案(Yjs + WebRTC),认为"first editor gets write lock"冲突解决足够。
|
||||
- 对方观点:Multi-Agent-System-Reliability认为大规模多Agent工作流需要显式协调基础设施。
|
||||
- 注:两者针对的场景不同(Nexus-Spatial侧重人类协作者的多人实时编辑;Multi-Agent-System-Reliability侧重自主Agent间的任务协调),但底层技术选择有潜在影响——大型Agent管道编排是否也需要CRDT风格的去中心化协调值得进一步研究。
|
||||
Reference in New Issue
Block a user