# dingdawg-agent-1 - Doramagic AI Context Pack

> 定位：安装前体验与判断资产。它帮助宿主 AI 有一个好的开始，但不代表已经安装、执行或验证目标项目。

## 充分原则

- **充分原则，不是压缩原则**：AI Context Pack 应该充分到让宿主 AI 在开工前理解项目价值、能力边界、使用入口、风险和证据来源；它可以分层组织，但不以最短摘要为目标。
- **压缩策略**：只压缩噪声和重复内容，不压缩会影响判断和开工质量的上下文。

## 给宿主 AI 的使用方式

你正在读取 Doramagic 为 dingdawg-agent-1 编译的 AI Context Pack。请把它当作开工前上下文：帮助用户理解适合谁、能做什么、如何开始、哪些必须安装后验证、风险在哪里。不要声称你已经安装、运行或执行了目标项目。

## Claim 消费规则

- **事实来源**：Repo Evidence + Claim/Evidence Graph；Human Wiki 只提供显著性、术语和叙事结构。
- **事实最低状态**：`supported`
- `supported`：可以作为项目事实使用，但回答中必须引用 claim_id 和证据路径。
- `weak`：只能作为低置信度线索，必须要求用户继续核实。
- `inferred`：只能用于风险提示或待确认问题，不能包装成项目事实。
- `unverified`：不得作为事实使用，应明确说证据不足。
- `contradicted`：必须展示冲突来源，不得替用户强行选择一个版本。

## 它最适合谁

- **想在安装前理解开源项目价值和边界的用户**：当前证据主要来自项目文档。 证据：`README.md` Claim：`clm_0019` supported 0.86

## 它能做什么

- **Agent Constitution Enforcement**（可做安装前预览）：YAML-defined behavioral contract evaluated before every action, with three enforcement tiers (hard_block, confirm_required, advisory). 证据：`README.md`, `ARCHITECTURE.md`, `docs/constitution-guide.md`, `docs/innovations.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Governance Gate Risk Assessment**（可做安装前预览）：Three-tier risk classification (PROCEED/REVIEW/HALT) applied to all requests and skill executions. 证据：`ARCHITECTURE.md`, `README.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **SHA-256 Hash-Chained Audit Trail**（可做安装前预览）：Tamper-evident audit logging with cryptographic hash chaining for all agent actions. 证据：`ARCHITECTURE.md`, `README.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Time-Lock Safety for Dangerous Actions**（需要安装后验证）：Mandatory cooling period (30-60s configurable) before executing dangerous operations, allowing user cancellation. 证据：`README.md`, `ARCHITECTURE.md`, `docs/innovations.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Multi-Provider LLM Brain with Fallback**（需要安装后验证）：Abstraction across OpenAI, Anthropic, and Google LLM providers with automatic fallback on provider failure. 证据：`README.md`, `docs/innovations.md`, `docs/comparison-vs-openclaw.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Pluggable Skill Architecture with Sandboxing**（需要安装后验证）：Self-contained skill packages with JSON manifests, sandboxed execution, and capability-based permissions. 证据：`ARCHITECTURE.md`, `docs/innovations.md`, `docs/skill-development.md`, `ARCHITECTURE.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Bridge-Gateway Messaging Architecture**（需要安装后验证）：Clean separation between messaging platforms and agent core via TypeScript bridges communicating via WebSocket RPC. 证据：`ARCHITECTURE.md`, `README.md`, `bridges/package.json`, `docs/innovations.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Trust Ledger & Reputation Tracking**（可做安装前预览）：Cryptographic reputation scoring for agent skills and actions with transparent ledger. 证据：`README.md`, `ARCHITECTURE.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **RBAC Access Control**（可做安装前预览）：Role-based access control with OWNER/OPERATOR/USER/PUBLIC roles for gateway authorization. 证据：`ARCHITECTURE.md`, `docs/security-model.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Filesystem & Execution Sandboxing**（需要安装后验证）：Workspace jailing with os.path.realpath(), Docker-based skill isolation, and resource limits. 证据：`ARCHITECTURE.md`, `docs/security-model.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Rate Limiting**（需要安装后验证）：Token bucket rate limiting for flood protection on all gateway endpoints. 证据：`ARCHITECTURE.md`, `docs/security-model.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **React Dashboard for Monitoring**（需要安装后验证）：Web-based monitoring and management interface for trust scores, audit trails, constitution viewing, and skill management. 证据：`ARCHITECTURE.md`, `dashboard/package.json`, `README.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Separation of Powers (Multi-Party Approval)**（需要安装后验证）：Critical actions require approval from independent agent or human before execution. 证据：`README.md`, `ARCHITECTURE.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Explain Mode (Decision Tracing)**（可做安装前预览）：Full decision trace with cryptographic proof linking to audit chain. 证据：`README.md`, `ARCHITECTURE.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **SQLite Database with WAL Mode**（需要安装后验证）：Concurrent-read database storing conversations, sessions, skills, audit trails, and trust ledger. 证据：`ARCHITECTURE.md`, `docs/architecture.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **CLI Terminal Interface**（可做安装前预览）：Command-line tool for interacting with agents via OAuth device flow or API key authentication. 证据：`gateway/packages/cli/README.md`, `gateway/packages/cli/package.json` Claim：`clm_0016` supported 0.86
- **WebSocket Real-Time Communication**（需要安装后验证）：WebSocket endpoint for real-time bridge-to-gateway communication with bearer token authentication. 证据：`docs/api-reference.md`, `ARCHITECTURE.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Convergence Guarantees (Bounded Execution)**（可做安装前预览）：Resource budgets and bounded execution loops to prevent infinite agent loops. 证据：`ARCHITECTURE.md`, `README.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等

## 怎么开始

- `git clone https://github.com/InnovativeSystemsGlobal/isg-agent-1.git` 证据：`README.md` Claim：`clm_0020` supported 0.86, `clm_0022` supported 0.86
- `pip install -r requirements.txt` 证据：`README.md` Claim：`clm_0021` supported 0.86
- `git clone https://github.com/innovative-systems-global/isg-agent-1.git` 证据：`docs/quickstart.md` Claim：`clm_0020` supported 0.86, `clm_0022` supported 0.86

## 继续前判断卡

- **当前建议**：先做角色匹配试用
- **为什么**：这个项目更像角色库，核心风险是选错角色或把角色文案当执行能力；先用 Prompt Preview 试角色匹配，再决定是否沙盒导入。

### 30 秒判断

- **现在怎么做**：先做角色匹配试用
- **最小安全下一步**：先用 Prompt Preview 试角色匹配；满意后再隔离导入
- **先别相信**：角色质量和任务匹配不能直接相信。
- **继续会触碰**：角色选择偏差、命令执行、本地环境或项目文件

### 现在可以相信

- **适合人群线索：想在安装前理解开源项目价值和边界的用户**（supported）：有 supported claim 或项目证据支撑，但仍不等于真实安装效果。 证据：`README.md` Claim：`clm_0019` supported 0.86
- **能力存在：Agent Constitution Enforcement**（supported）：可以相信项目包含这类能力线索；是否适合你的具体任务仍要试用或安装后验证。 证据：`README.md`, `ARCHITECTURE.md`, `docs/constitution-guide.md`, `docs/innovations.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86
- **能力存在：Governance Gate Risk Assessment**（supported）：可以相信项目包含这类能力线索；是否适合你的具体任务仍要试用或安装后验证。 证据：`ARCHITECTURE.md`, `README.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86
- **能力存在：SHA-256 Hash-Chained Audit Trail**（supported）：可以相信项目包含这类能力线索；是否适合你的具体任务仍要试用或安装后验证。 证据：`ARCHITECTURE.md`, `README.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86
- **能力存在：Time-Lock Safety for Dangerous Actions**（supported）：可以相信项目包含这类能力线索；是否适合你的具体任务仍要试用或安装后验证。 证据：`README.md`, `ARCHITECTURE.md`, `docs/innovations.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86
- **能力存在：Multi-Provider LLM Brain with Fallback**（supported）：可以相信项目包含这类能力线索；是否适合你的具体任务仍要试用或安装后验证。 证据：`README.md`, `docs/innovations.md`, `docs/comparison-vs-openclaw.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86

### 现在还不能相信

- **角色质量和任务匹配不能直接相信。**（unverified）：角色库证明有很多角色，不证明每个角色都适合你的具体任务，也不证明角色能产生高质量结果。
- **不能把角色文案当成真实执行能力。**（unverified）：安装前只能判断角色描述和任务画像是否匹配，不能证明它能在宿主 AI 里完成任务。
- **真实输出质量不能在安装前相信。**（unverified）：Prompt Preview 只能展示引导方式，不能证明真实项目中的结果质量。
- **宿主 AI 版本兼容性不能在安装前相信。**（unverified）：Claude、Cursor、Codex、Gemini 等宿主加载规则和版本差异必须在真实环境验证。
- **不会污染现有宿主 AI 行为，不能直接相信。**（inferred）：Skill、plugin、AGENTS/CLAUDE/GEMINI 指令可能改变宿主 AI 的默认行为。
- **可安全回滚不能默认相信。**（unverified）：除非项目明确提供卸载和恢复说明，否则必须先在隔离环境验证。
- **真实安装后是否与用户当前宿主 AI 版本兼容？**（unverified）：兼容性只能通过实际宿主环境验证。
- **项目输出质量是否满足用户具体任务？**（unverified）：安装前预览只能展示流程和边界，不能替代真实评测。

### 继续会触碰什么

- **角色选择偏差**：用户对任务应该由哪个专家角色处理的判断。 原因：选错角色会让 AI 从错误专业视角回答，浪费时间或误导决策。
- **命令执行**：包管理器、网络下载、本地插件目录、项目配置或用户主目录。 原因：运行第一条命令就可能产生环境改动；必须先判断是否值得跑。 证据：`README.md`, `docs/quickstart.md`
- **本地环境或项目文件**：安装结果、插件缓存、项目配置或本地依赖目录。 原因：安装前无法证明写入范围和回滚方式，需要隔离验证。 证据：`ARCHITECTURE.md`, `README.md`, `bridges/package.json`, `dashboard/package.json` 等
- **宿主 AI 上下文**：AI Context Pack、Prompt Preview、Skill 路由、风险规则和项目事实。 原因：导入上下文会影响宿主 AI 后续判断，必须避免把未验证项包装成事实。

### 最小安全下一步

- **先跑 Prompt Preview**：先用交互式试用验证任务画像和角色匹配，不要先导入整套角色库。（适用：任何项目都适用，尤其是输出质量未知时。）
- **只在隔离目录或测试账号试装**：避免安装命令污染主力宿主 AI、真实项目或用户主目录。（适用：存在命令执行、插件配置或本地写入线索时。）
- **安装后只验证一个最小任务**：先验证加载、兼容、输出质量和回滚，再决定是否深用。（适用：准备从试用进入真实工作流时。）

### 退出方式

- **保留安装前状态**：记录原始宿主配置和项目状态，后续才能判断是否可恢复。
- **保留原始角色选择记录**：如果输出偏题，可以回到任务画像阶段重新选择角色，而不是继续沿着错误角色推进。
- **记录安装命令和写入路径**：没有明确卸载说明时，至少要知道哪些目录或配置需要手动清理。
- **如果没有回滚路径，不进入主力环境**：不可回滚是继续前阻断项，不应靠信任或运气继续。

## 哪些只能预览

- 解释项目适合谁和能做什么
- 基于项目文档演示典型对话流程
- 帮助用户判断是否值得安装或继续研究

## 哪些必须安装后验证

- 真实安装 Skill、插件或 CLI
- 执行脚本、修改本地文件或访问外部服务
- 验证真实输出质量、性能和兼容性

## 边界与风险判断卡

- **把安装前预览误认为真实运行**：用户可能高估项目已经完成的配置、权限和兼容性验证。 处理方式：明确区分 prompt_preview_can_do 与 runtime_required。 Claim：`clm_0023` inferred 0.45
- **命令执行会修改本地环境**：安装命令可能写入用户主目录、宿主插件目录或项目配置。 处理方式：先在隔离环境或测试账号中运行。 证据：`README.md`, `docs/quickstart.md` Claim：`clm_0024` supported 0.86
- **风险**： 处理方式：
- **风险**： 处理方式：
- **风险**： 处理方式：
- **风险**： 处理方式：
- **待确认**：真实安装后是否与用户当前宿主 AI 版本兼容？。原因：兼容性只能通过实际宿主环境验证。
- **待确认**：项目输出质量是否满足用户具体任务？。原因：安装前预览只能展示流程和边界，不能替代真实评测。
- **待确认**：安装命令是否需要网络、权限或全局写入？。原因：这影响企业环境和个人环境的安装风险。

## 开工前工作上下文

### 加载顺序

- 先读取 how_to_use.host_ai_instruction，建立安装前判断资产的边界。
- 读取 claim_graph_summary，确认事实来自 Claim/Evidence Graph，而不是 Human Wiki 叙事。
- 再读取 intended_users、capabilities 和 quick_start_candidates，判断用户是否匹配。
- 需要执行具体任务时，优先查 role_skill_index，再查 evidence_index。
- 遇到真实安装、文件修改、网络访问、性能或兼容性问题时，转入 risk_card 和 boundaries.runtime_required。

### 任务路由

- **Agent Constitution Enforcement**：先基于 role_skill_index / evidence_index 帮用户挑选可用角色、Skill 或工作流。 边界：可做安装前 Prompt 体验。 证据：`README.md`, `ARCHITECTURE.md`, `docs/constitution-guide.md`, `docs/innovations.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Governance Gate Risk Assessment**：先基于 role_skill_index / evidence_index 帮用户挑选可用角色、Skill 或工作流。 边界：可做安装前 Prompt 体验。 证据：`ARCHITECTURE.md`, `README.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **SHA-256 Hash-Chained Audit Trail**：先基于 role_skill_index / evidence_index 帮用户挑选可用角色、Skill 或工作流。 边界：可做安装前 Prompt 体验。 证据：`ARCHITECTURE.md`, `README.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Time-Lock Safety for Dangerous Actions**：先说明这是安装后验证能力，再给出安装前检查清单。 边界：必须真实安装或运行后验证。 证据：`README.md`, `ARCHITECTURE.md`, `docs/innovations.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Multi-Provider LLM Brain with Fallback**：先说明这是安装后验证能力，再给出安装前检查清单。 边界：必须真实安装或运行后验证。 证据：`README.md`, `docs/innovations.md`, `docs/comparison-vs-openclaw.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Pluggable Skill Architecture with Sandboxing**：先说明这是安装后验证能力，再给出安装前检查清单。 边界：必须真实安装或运行后验证。 证据：`ARCHITECTURE.md`, `docs/innovations.md`, `docs/skill-development.md`, `ARCHITECTURE.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Bridge-Gateway Messaging Architecture**：先说明这是安装后验证能力，再给出安装前检查清单。 边界：必须真实安装或运行后验证。 证据：`ARCHITECTURE.md`, `README.md`, `bridges/package.json`, `docs/innovations.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Trust Ledger & Reputation Tracking**：先基于 role_skill_index / evidence_index 帮用户挑选可用角色、Skill 或工作流。 边界：可做安装前 Prompt 体验。 证据：`README.md`, `ARCHITECTURE.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **RBAC Access Control**：先基于 role_skill_index / evidence_index 帮用户挑选可用角色、Skill 或工作流。 边界：可做安装前 Prompt 体验。 证据：`ARCHITECTURE.md`, `docs/security-model.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Filesystem & Execution Sandboxing**：先说明这是安装后验证能力，再给出安装前检查清单。 边界：必须真实安装或运行后验证。 证据：`ARCHITECTURE.md`, `docs/security-model.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Rate Limiting**：先说明这是安装后验证能力，再给出安装前检查清单。 边界：必须真实安装或运行后验证。 证据：`ARCHITECTURE.md`, `docs/security-model.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **React Dashboard for Monitoring**：先说明这是安装后验证能力，再给出安装前检查清单。 边界：必须真实安装或运行后验证。 证据：`ARCHITECTURE.md`, `dashboard/package.json`, `README.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Separation of Powers (Multi-Party Approval)**：先说明这是安装后验证能力，再给出安装前检查清单。 边界：必须真实安装或运行后验证。 证据：`README.md`, `ARCHITECTURE.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Explain Mode (Decision Tracing)**：先基于 role_skill_index / evidence_index 帮用户挑选可用角色、Skill 或工作流。 边界：可做安装前 Prompt 体验。 证据：`README.md`, `ARCHITECTURE.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **SQLite Database with WAL Mode**：先说明这是安装后验证能力，再给出安装前检查清单。 边界：必须真实安装或运行后验证。 证据：`ARCHITECTURE.md`, `docs/architecture.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **CLI Terminal Interface**：先基于 role_skill_index / evidence_index 帮用户挑选可用角色、Skill 或工作流。 边界：可做安装前 Prompt 体验。 证据：`gateway/packages/cli/README.md`, `gateway/packages/cli/package.json` Claim：`clm_0016` supported 0.86
- **WebSocket Real-Time Communication**：先说明这是安装后验证能力，再给出安装前检查清单。 边界：必须真实安装或运行后验证。 证据：`docs/api-reference.md`, `ARCHITECTURE.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等
- **Convergence Guarantees (Bounded Execution)**：先基于 role_skill_index / evidence_index 帮用户挑选可用角色、Skill 或工作流。 边界：可做安装前 Prompt 体验。 证据：`ARCHITECTURE.md`, `README.md` Claim：`clm_0001` supported 0.86, `clm_0002` supported 0.86, `clm_0003` supported 0.86, `clm_0004` supported 0.86 等

### 上下文规模

- 文件总数：671
- 重要文件覆盖：40/671
- 证据索引条目：80
- 角色 / Skill 条目：21

### 证据不足时的处理

- **missing_evidence**：说明证据不足，要求用户提供目标文件、README 段落或安装后验证记录；不要补全事实。
- **out_of_scope_request**：说明该任务超出当前 AI Context Pack 证据范围，并建议用户先查看 Human Manual 或真实安装后验证。
- **runtime_request**：给出安装前检查清单和命令来源，但不要替用户执行命令或声称已执行。
- **source_conflict**：同时展示冲突来源，标记为待核实，不要强行选择一个版本。

## Prompt Recipes

### 适配判断

- 目标：判断这个项目是否适合用户当前任务。
- 预期输出：适配结论、关键理由、证据引用、安装前可预览内容、必须安装后验证内容、下一步建议。

```text
请基于 dingdawg-agent-1 的 AI Context Pack，先问我 3 个必要问题，然后判断它是否适合我的任务。回答必须包含：适合谁、能做什么、不能做什么、是否值得安装、证据来自哪里。所有项目事实必须引用 evidence_refs、source_paths 或 claim_id。
```

### 安装前体验

- 目标：让用户在安装前感受核心工作流，同时避免把预览包装成真实能力或营销承诺。
- 预期输出：一段带边界标签的体验剧本、安装后验证清单和谨慎建议；不含真实运行承诺或强营销表述。

```text
请把 dingdawg-agent-1 当作安装前体验资产，而不是已安装工具或真实运行环境。

请严格输出四段：
1. 先问我 3 个必要问题。
2. 给出一段“体验剧本”：用 [安装前可预览]、[必须安装后验证]、[证据不足] 三种标签展示它可能如何引导工作流。
3. 给出安装后验证清单：列出哪些能力只有真实安装、真实宿主加载、真实项目运行后才能确认。
4. 给出谨慎建议：只能说“值得继续研究/试装”“先补充信息后再判断”或“不建议继续”，不得替项目背书。

硬性边界：
- 不要声称已经安装、运行、执行测试、修改文件或产生真实结果。
- 不要写“自动适配”“确保通过”“完美适配”“强烈建议安装”等承诺性表达。
- 如果描述安装后的工作方式，必须使用“如果安装成功且宿主正确加载 Skill，它可能会……”这种条件句。
- 体验剧本只能写成“示例台词/假设流程”：使用“可能会询问/可能会建议/可能会展示”，不要写“已写入、已生成、已通过、正在运行、正在生成”。
- Prompt Preview 不负责给安装命令；如用户准备试装，只能提示先阅读 Quick Start 和 Risk Card，并在隔离环境验证。
- 所有项目事实必须来自 supported claim、evidence_refs 或 source_paths；inferred/unverified 只能作风险或待确认项。

```

### 角色 / Skill 选择

- 目标：从项目里的角色或 Skill 中挑选最匹配的资产。
- 预期输出：候选角色或 Skill 列表，每项包含适用场景、证据路径、风险边界和是否需要安装后验证。

```text
请读取 role_skill_index，根据我的目标任务推荐 3-5 个最相关的角色或 Skill。每个推荐都要说明适用场景、可能输出、风险边界和 evidence_refs。
```

### 风险预检

- 目标：安装或引入前识别环境、权限、规则冲突和质量风险。
- 预期输出：环境、权限、依赖、许可、宿主冲突、质量风险和未知项的检查清单。

```text
请基于 risk_card、boundaries 和 quick_start_candidates，给我一份安装前风险预检清单。不要替我执行命令，只说明我应该检查什么、为什么检查、失败会有什么影响。
```

### 宿主 AI 开工指令

- 目标：把项目上下文转成一次对话开始前的宿主 AI 指令。
- 预期输出：一段边界明确、证据引用明确、适合复制给宿主 AI 的开工前指令。

```text
请基于 dingdawg-agent-1 的 AI Context Pack，生成一段我可以粘贴给宿主 AI 的开工前指令。这段指令必须遵守 not_runtime=true，不能声称项目已经安装、运行或产生真实结果。
```


## 角色 / Skill 索引

- 共索引 21 个角色 / Skill / 项目文档条目。

- **ISG Agent 1**（project_doc）：! CI https://github.com/InnovativeSystemsGlobal/isg-agent-1/actions/workflows/ci.yml/badge.svg https://github.com/InnovativeSystemsGlobal/isg-agent-1/actions/workflows/ci.yml ! License: MIT https://img.shields.io/badge/License-MIT-blue.svg LICENSE ! Python 3.11+ https://img.shields.io/badge/python-3.11+-blue.svg https://www.python.org/downloads/ 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`README.md`
- **Architecture Overview**（project_doc）：ISG Agent 1 is composed of three main services: 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/architecture.md`
- **Getting Started**（project_doc）：This is a Next.js https://nextjs.org project bootstrapped with create-next-app https://nextjs.org/docs/app/api-reference/cli/create-next-app . 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`gateway/frontend/README.md`
- **@dingdawg/cli**（project_doc）：DingDawg Agent CLI — Talk to your AI agents from the terminal. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`gateway/packages/cli/README.md`
- **Contributing to ISG Agent 1**（project_doc）：Thank you for your interest in contributing. ISG Agent 1 is a security-hardened agent platform, and contributions must meet our quality and security standards. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`CONTRIBUTING.md`
- **Agent 1 Checkout Load Test**（project_doc）：Step Endpoint Method Notes ------ ---------- -------- ------- 1 /api/v1/live GET Liveness baseline per VU 2 /auth/login POST Per-VU JWT token acquisition 3 /api/v1/payments/create-checkout-session POST Primary load target 4 /api/v1/payments/status GET Read-only subscription status 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`gateway/tests/load/README.md`
- **ISG Agent 1 -- Architecture Overview**（project_doc）：ISG Agent 1 -- Architecture Overview 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`ARCHITECTURE.md`
- **API Reference**（project_doc）：Method Path Description -------- ----------- -------------------------- GET /health Gateway health check GET /docs Interactive API docs GET /redoc Alternative API docs 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/api-reference.md`
- **ISG Agent 1 vs. Open-Source Alternatives**（project_doc）：ISG Agent 1 vs. Open-Source Alternatives 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/comparison-vs-openclaw.md`
- **Constitution Guide**（project_doc）：The constitution is a YAML file that defines the behavioral rules your agent must follow. It is evaluated before every action the agent takes, providing a governance layer between the user's request and the agent's execution. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/constitution-guide.md`
- **Deployment Guide**（project_doc）：Copy .env.example to .env and fill in all required values before running Docker Compose. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/deployment.md`
- **ISG Agent 1 Innovations**（project_doc）：A YAML-defined behavioral contract that governs all agent actions. The constitution is evaluated before every action and cannot be overridden by user instructions or LLM outputs. 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/innovations.md`
- **Quickstart Guide**（project_doc）：- Python 3.11+ - Node.js 20+ - Docker optional, for containerized deployment 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/quickstart.md`
- **Security Model**（project_doc）：ISG Agent 1 implements multiple security layers: 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/security-model.md`
- **Skill Development Guide**（project_doc）：Each skill is a directory under skills/ containing: 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`docs/skill-development.md`
- **DingDawg Main -- Done Definition Press Release Format**（project_doc）：DingDawg Main -- Done Definition Press Release Format 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`gateway/docs/DINGDAWG_MAIN_DONE_DEFINITION.md`
- **DingDawg Agent 1 -- Done Definition Press Release Format**（project_doc）：DingDawg Agent 1 -- Done Definition Press Release Format 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`gateway/docs/PRODUCT_DONE_DEFINITION.md`
- **Project School -- Done Definition Press Release Format**（project_doc）：Project School -- Done Definition Press Release Format 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`gateway/docs/PROJECT_SCHOOL_DONE_DEFINITION.md`
- **Competitive Analysis: AI Agent Skills / Actions / Tools**（project_doc）：Competitive Analysis: AI Agent Skills / Actions / Tools How the Top Platforms Let Agents DO Things Not Just Chat 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`gateway/docs/RESEARCH_AGENT_SKILLS_COMPETITIVE_ANALYSIS.md`
- **Pre-Deploy Smoke Test Checklist**（project_doc）：Purpose: Universal smoke test that applies to ALL products before any deployment claim. Rule: If any item fails, the product is NOT ready to deploy. No exceptions. No "we'll fix it after launch." 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`gateway/docs/SMOKE_TEST_CHECKLIST.md`
- **STRUCTURAL PRINCIPLES BREAKDOWN**（project_doc）：STRUCTURAL PRINCIPLES BREAKDOWN Every Principle Converted to Executable DO/DON'T/MEASURE Rules Date: 2026-02-19 Source: Session 12 Research 4 documents Status: REFERENCE DOCUMENT -- Use for gate/hook/check implementation 激活提示：当用户需要理解项目结构、安装方式或边界时参考。 证据：`gateway/docs/STRUCTURAL_PRINCIPLES_BREAKDOWN.md`

## 证据索引

- 共索引 80 条证据。

- **ISG Agent 1**（documentation）：! CI https://github.com/InnovativeSystemsGlobal/isg-agent-1/actions/workflows/ci.yml/badge.svg https://github.com/InnovativeSystemsGlobal/isg-agent-1/actions/workflows/ci.yml ! License: MIT https://img.shields.io/badge/License-MIT-blue.svg LICENSE ! Python 3.11+ https://img.shields.io/badge/python-3.11+-blue.svg https://www.python.org/downloads/ 证据：`README.md`
- **Architecture Overview**（documentation）：ISG Agent 1 is composed of three main services: 证据：`docs/architecture.md`
- **Getting Started**（documentation）：This is a Next.js https://nextjs.org project bootstrapped with create-next-app https://nextjs.org/docs/app/api-reference/cli/create-next-app . 证据：`gateway/frontend/README.md`
- **@dingdawg/cli**（documentation）：DingDawg Agent CLI — Talk to your AI agents from the terminal. 证据：`gateway/packages/cli/README.md`
- **Package**（package_manifest）：{ "name": "@isg-agent/bridges", "version": "0.1.0", "description": "ISG Agent 1 messaging platform bridges Discord, Telegram, WebSocket ", "private": true, "type": "module", "main": "dist/index.js", "types": "dist/index.d.ts", "scripts": { "build": "tsc", "dev": "tsc --watch", "test": "vitest run", "test:watch": "vitest", "lint": "tsc --noEmit", "clean": "rm -rf dist" }, "dependencies": { "discord.js": "^14.16.0", "telegraf": "^4.16.0", "ws": "^8.18.0", "zod": "^3.23.0" }, "devDependencies": { "@types/node": "^22.10.0", "@types/ws": "^8.5.13", "typescript": "^5.7.0", "vitest": "^2.1.0" }, "engines": { "node": " =20.0.0" }, "license": "MIT" } 证据：`bridges/package.json`
- **Package**（package_manifest）：{ "name": "@isg-agent/dashboard", "version": "0.1.0", "description": "ISG Agent 1 monitoring and management dashboard", "private": true, "type": "module", "scripts": { "dev": "vite", "build": "tsc && vite build", "preview": "vite preview", "test": "vitest run", "test:watch": "vitest", "lint": "tsc --noEmit" }, "dependencies": { "react": "^18.3.1", "react-dom": "^18.3.1" }, "devDependencies": { "@types/react": "^18.3.12", "@types/react-dom": "^18.3.1", "@vitejs/plugin-react": "^4.3.4", "typescript": "^5.7.0", "vite": "^6.0.0", "vitest": "^2.1.0" }, "engines": { "node": " =20.0.0" }, "license": "MIT" } 证据：`dashboard/package.json`
- **Package**（package_manifest）：{ "private": true, "workspaces": "bridges", "dashboard" , "scripts": { "build:bridges": "npm -w bridges run build", "build:dashboard": "npm -w dashboard run build", "test:bridges": "npm -w bridges run test", "test:dashboard": "npm -w dashboard run test" } } 证据：`package.json`
- **Contributing to ISG Agent 1**（documentation）：Thank you for your interest in contributing. ISG Agent 1 is a security-hardened agent platform, and contributions must meet our quality and security standards. 证据：`CONTRIBUTING.md`
- **Package**（package_manifest）：{ "name": "frontend", "version": "0.1.0", "license": "UNLICENSED", "private": true, "scripts": { "dev": "next dev", "build": "next build", "start": "next start", "lint": "eslint", "test": "vitest run", "test:watch": "vitest", "test:coverage": "vitest run --coverage", "test:e2e": "playwright test --project=production", "test:e2e:local": "playwright test --project=local", "test:e2e:vr": "playwright test e2e/visual-regression.spec.ts --project=local", "test:e2e:vr:update": "playwright test e2e/visual-regression.spec.ts --project=local --update-snapshots", "test:e2e:smoke": "playwright test e2e/protected-pages-smoke.spec.ts --project=local", "test:e2e:mobile": "playwright test --project=mobile-… 证据：`gateway/frontend/package.json`
- **Package**（package_manifest）：{ "name": "@dingdawg/chat-ui", "version": "0.1.0", "description": "DingDawg shared SMS-like chat UI components", "author": "Innovative Systems Global LLC", "license": "MIT", "type": "module", "main": "./src/index.ts", "types": "./src/index.ts", "exports": { ".": "./src/index.ts" }, "keywords": "chat", "ui", "sms", "dingdawg", "react" , "peerDependencies": { "react": " =19.0.0", "react-dom": " =19.0.0", "react-virtuoso": " =4.0.0", "lucide-react": " =0.5.0", "clsx": " =2.0.0", "tailwind-merge": " =3.0.0" }, "devDependencies": { "@types/react": "^19", "@types/react-dom": "^19", "typescript": "^5", "@testing-library/react": "^16.0.0", "@testing-library/jest-dom": "^6.0.0", "@testing-library/us… 证据：`gateway/packages/chat-ui/package.json`
- **Package**（package_manifest）：{ "name": "@dingdawg/cli", "version": "0.1.0", "description": "DingDawg Agent CLI — Talk to your AI agents from the terminal", "bin": { "dd": "./bin/dd.js" }, "type": "module", "engines": { "node": " =18" }, "keywords": "ai", "agent", "cli", "dingdawg", "llm" , "license": "MIT", "author": "Innovative Systems Global LLC", "repository": { "type": "git", "url": "https://github.com/dingdawg/cli" }, "files": "bin/", "src/", "README.md" } 证据：`gateway/packages/cli/package.json`
- **Package**（package_manifest）：{ "name": "@dingdawg/sdk", "version": "2.0.2", "description": "DingDawg Agent SDK \u2014 durable AI agents with crash-proof exactly-once execution + IPFS checkpoints", "main": "./dist/index.cjs", "module": "./dist/index.js", "types": "./dist/index.d.ts", "exports": { ".": { "import": "./dist/index.js", "require": "./dist/index.cjs", "types": "./dist/index.d.ts" } }, "engines": { "node": " =18" }, "scripts": { "build": "tsc && tsc -p tsconfig.cjs.json && node scripts/postbuild.js", "test": "node --experimental-vm-modules node modules/.bin/jest --testPathPattern='src/ tests '", "typecheck": "tsc --noEmit" }, "keywords": "ai", "agent", "sdk", "dingdawg", "mcp", "a2a", "llm", "b2b2c" , "license… 证据：`gateway/packages/sdk/package.json`
- **Agent 1 Checkout Load Test**（documentation）：Step Endpoint Method Notes ------ ---------- -------- ------- 1 /api/v1/live GET Liveness baseline per VU 2 /auth/login POST Per-VU JWT token acquisition 3 /api/v1/payments/create-checkout-session POST Primary load target 4 /api/v1/payments/status GET Read-only subscription status 证据：`gateway/tests/load/README.md`
- **License**（source_file）：Copyright c 2026 Innovative Systems Global LLC 证据：`LICENSE`
- **ISG Agent 1 -- Architecture Overview**（documentation）：ISG Agent 1 -- Architecture Overview 证据：`ARCHITECTURE.md`
- **API Reference**（documentation）：Method Path Description -------- ----------- -------------------------- GET /health Gateway health check GET /docs Interactive API docs GET /redoc Alternative API docs 证据：`docs/api-reference.md`
- **ISG Agent 1 vs. Open-Source Alternatives**（documentation）：ISG Agent 1 vs. Open-Source Alternatives 证据：`docs/comparison-vs-openclaw.md`
- **Constitution Guide**（documentation）：The constitution is a YAML file that defines the behavioral rules your agent must follow. It is evaluated before every action the agent takes, providing a governance layer between the user's request and the agent's execution. 证据：`docs/constitution-guide.md`
- **Deployment Guide**（documentation）：Copy .env.example to .env and fill in all required values before running Docker Compose. 证据：`docs/deployment.md`
- **ISG Agent 1 Innovations**（documentation）：A YAML-defined behavioral contract that governs all agent actions. The constitution is evaluated before every action and cannot be overridden by user instructions or LLM outputs. 证据：`docs/innovations.md`
- **Quickstart Guide**（documentation）：- Python 3.11+ - Node.js 20+ - Docker optional, for containerized deployment 证据：`docs/quickstart.md`
- **Security Model**（documentation）：ISG Agent 1 implements multiple security layers: 证据：`docs/security-model.md`
- **Skill Development Guide**（documentation）：Each skill is a directory under skills/ containing: 证据：`docs/skill-development.md`
- **DingDawg Main -- Done Definition Press Release Format**（documentation）：DingDawg Main -- Done Definition Press Release Format 证据：`gateway/docs/DINGDAWG_MAIN_DONE_DEFINITION.md`
- **DingDawg Agent 1 -- Done Definition Press Release Format**（documentation）：DingDawg Agent 1 -- Done Definition Press Release Format 证据：`gateway/docs/PRODUCT_DONE_DEFINITION.md`
- **Project School -- Done Definition Press Release Format**（documentation）：Project School -- Done Definition Press Release Format 证据：`gateway/docs/PROJECT_SCHOOL_DONE_DEFINITION.md`
- **Competitive Analysis: AI Agent Skills / Actions / Tools**（documentation）：Competitive Analysis: AI Agent Skills / Actions / Tools How the Top Platforms Let Agents DO Things Not Just Chat 证据：`gateway/docs/RESEARCH_AGENT_SKILLS_COMPETITIVE_ANALYSIS.md`
- **Pre-Deploy Smoke Test Checklist**（documentation）：Purpose: Universal smoke test that applies to ALL products before any deployment claim. Rule: If any item fails, the product is NOT ready to deploy. No exceptions. No "we'll fix it after launch." 证据：`gateway/docs/SMOKE_TEST_CHECKLIST.md`
- **STRUCTURAL PRINCIPLES BREAKDOWN**（documentation）：STRUCTURAL PRINCIPLES BREAKDOWN Every Principle Converted to Executable DO/DON'T/MEASURE Rules Date: 2026-02-19 Source: Session 12 Research 4 documents Status: REFERENCE DOCUMENT -- Use for gate/hook/check implementation 证据：`gateway/docs/STRUCTURAL_PRINCIPLES_BREAKDOWN.md`
- **Index**（source_file）：ISG Agent 1 Dashboard 证据：`dashboard/index.html`
- **Instructions**（source_file）：Playwright Test Report var NA=Object.defineProperty;var BA= l,u,c = u in l?NA l,u,{enumerable:!0,configurable:!0,writable:!0,value:c} :l u =c;var yn= l,u,c = BA l,typeof u!="symbol"?u+"":u,c ; function {const u=document.createElement "link" .relList;if u&&u.supports&&u.supports "modulepreload" return;for const r of document.querySelectorAll 'link rel="modulepreload" ' f r ;new MutationObserver r= {for const o of r if o.type==="childList" for const h of o.addedNodes h.tagName==="LINK"&&h.rel==="modulepreload"&&f h } .observe document,{childList:!0,subtree:!0} ;function c r {const o={};return r.integrity&& o.integrity=r.integrity ,r.referrerPolicy&& o.referrerPolicy=r.referrerPolicy ,r.crossO… 证据：`gateway/frontend/playwright-report/index.html`
- **Generate Screenshots**（source_file）：async function tryProductionScreenshot browser, url, viewport, outputPath ⋮---- async function takeFallbackScreenshot browser, html, viewport, outputPath ⋮---- async function main 证据：`gateway/frontend/scripts/generate-screenshots.js`
- **Page**（source_file）：import { useCallback, useEffect, useState } from "react"; import { PieChart, Pie, Cell, Tooltip, ResponsiveContainer, Legend, } from "recharts"; import StatCard from "@/components/admin/StatCard"; import DataTable from "@/components/admin/DataTable"; import type { ColumnDef } from "@/components/admin/DataTable"; import StatusDot from "@/components/admin/StatusDot"; import { usePolling } from "@/hooks/usePolling"; ⋮---- import type { AdminAgent, TemplateDistributionItem } from "@/services/api/adminService"; import { useAdminAgentsStore } from "@/store/adminAgentsStore"; ⋮---- type StatusFilter = typeof STATUS FILTERS number ; ⋮---- function agentStatusColor status: AdminAgent "status" : "gre… 证据：`gateway/frontend/src/app/admin/agents/page.tsx`
- **Page**（source_file）：import { useState, useEffect, useCallback, useRef } from "react"; import { Bell, BellOff, AlertTriangle, Info, XCircle, CheckCircle2, ChevronDown, ChevronUp, RefreshCw, AlertCircle, Shield, CreditCard, Settings2, Plug, Activity, } from "lucide-react"; import { cn } from "@/lib/utils"; import { configureAlerts, type Alert, type AlertSeverity, type AlertSource, type AlertConfig, } from "@/services/api/adminService"; import { useAdminAlertsStore, type AlertFilter } from "@/store/adminAlertsStore"; ⋮---- function formatRelativeTime timestamp: string : string ⋮---- function AlertBadge ⋮---- className= ⋮---- function SourceIcon ⋮---- interface AlertCardProps { alert: Alert; onAcknowledge: id: str… 证据：`gateway/frontend/src/app/admin/alerts/page.tsx`
- **Page**（source_file）：import { useCallback, useState } from "react"; import { BarChart, Bar, XAxis, YAxis, Tooltip, ResponsiveContainer, } from "recharts"; import { Users, AlertCircle, RefreshCw, TrendingDown, Activity, Search, } from "lucide-react"; import { cn } from "@/lib/utils"; import { useAdminRevenueStore } from "@/store/adminRevenueStore"; import { usePolling } from "@/hooks/usePolling"; import AdminRoute from "@/components/auth/AdminRoute"; import StatusDot from "@/components/admin/StatusDot"; import DataTable, { type ColumnDef } from "@/components/admin/DataTable"; import type { Contact, FunnelData } from "@/store/adminRevenueStore"; ⋮---- interface FunnelBarProps { label: string; value: number; maxVa… 证据：`gateway/frontend/src/app/admin/crm/page.tsx`
- **Page**（source_file）：import { useCallback, useRef, useState } from "react"; import { BarChart, Bar, XAxis, YAxis, Tooltip, ResponsiveContainer, Cell, } from "recharts"; import { usePolling } from "@/hooks/usePolling"; ⋮---- import type { ErrorEntry, HealthDetailed, EndpointStat, ClientErrorEntry, } from "@/services/api/adminService"; ⋮---- function formatUptime seconds: number : string ⋮---- function formatTime isoString: string : string ⋮---- function ErrorCard ⋮---- function StatPill { label, value, warn, }: { label: string; value: string; warn?: boolean; } ⋮---- function EndpointRow ⋮---- onClick= ⋮---- async function fetchAll ⋮---- async function handleClear 证据：`gateway/frontend/src/app/admin/debug/page.tsx`
- **Page**（source_file）：import { useEffect, useState, useCallback } from "react"; import { Rocket, RefreshCw, AlertCircle, CheckCircle2, Clock, ChevronDown, Megaphone, Zap, ShoppingBag, Calendar, BarChart3, MessageSquare, Users, Globe, } from "lucide-react"; import { cn } from "@/lib/utils"; import { getAdminTemplates, getMarketingAgentStatus, deployMarketingAgent, deployAgent, getDeploymentHistory, type AdminTemplate, type MarketingAgentStatus, type DeploymentRecord, type DeploymentStatus, } from "@/services/api/adminService"; ⋮---- function formatTimestamp iso: string : string ⋮---- function getSectorIcon sector: string : React.ComponentType ⋮---- <div className= ⋮---- // ─── Quick Deploy ───────────────────────… 证据：`gateway/frontend/src/app/admin/deploy/page.tsx`
- **Page**（source_file）：import { useCallback, useState } from "react"; import { BarChart, Bar, XAxis, YAxis, Tooltip, ResponsiveContainer, Legend, } from "recharts"; import StatusDot from "@/components/admin/StatusDot"; import { usePolling } from "@/hooks/usePolling"; ⋮---- import type { IntegrationHealth, IntegrationStatus, IntegrationTestResult, } from "@/services/api/adminService"; ⋮---- interface IntegrationMeta { key: string; name: string; icon: string; hasWebhook: boolean; } ⋮---- function statusColor s: IntegrationStatus : "green" "red" "gray" ⋮---- function statusLabel s: IntegrationStatus : string ⋮---- function formatTime isoString: string null : string ⋮---- function RateBar ⋮---- <StatusDot color= 证据：`gateway/frontend/src/app/admin/integrations/page.tsx`
- **Layout**（source_file）：import type { Metadata } from "next"; import AdminRoute from "@/components/auth/AdminRoute"; import AdminShell from "@/components/layout/AdminShell"; import { AdminErrorBoundary } from "@/components/admin/AdminErrorBoundary"; ⋮---- export default function AdminLayout { children, }: { children: React.ReactNode; } 证据：`gateway/frontend/src/app/admin/layout.tsx`
- **Page**（source_file）：import { useState, useEffect, useCallback, useRef } from "react"; import { Mail, BarChart2, MousePointerClick, AlertCircle, RefreshCw, Megaphone, Rocket, CheckCircle2, Loader2, TrendingUp, Send, } from "lucide-react"; import { cn } from "@/lib/utils"; import { getCampaigns, getEmailStats, deployMarketingAgent, type Campaign, type CampaignStatus, type EmailStats, } from "@/services/api/adminService"; ⋮---- function formatPct value: number : string ⋮---- function formatDate iso: string : string ⋮---- function formatNumber n: number : string ⋮---- className= ⋮---- ; accent?: boolean; } ⋮---- onClick= 证据：`gateway/frontend/src/app/admin/marketing/page.tsx`
- **Page**（source_file）：import { useCallback, useEffect } from "react"; import { useAdminStore } from "@/store/adminStore"; import { usePolling } from "@/hooks/usePolling"; import StatCard from "@/components/admin/StatCard"; import AlertBadge from "@/components/admin/AlertBadge"; import StatusDot from "@/components/admin/StatusDot"; import { getAlerts, getHealthDetailed } from "@/services/api/adminService"; import { useState } from "react"; import type { Alert, HealthDetailed } from "@/services/api/adminService"; import { cn } from "@/lib/utils"; ⋮---- function SkeletonCard ⋮---- function formatUptime seconds: number : string ⋮---- className= ⋮---- 证据：`gateway/frontend/src/app/admin/page.tsx`
- **Page**（source_file）：import { useCallback, useEffect, useState } from "react"; import { AreaChart, Area, XAxis, YAxis, Tooltip, ResponsiveContainer, } from "recharts"; import { DollarSign, Users, TrendingUp, BarChart3, AlertCircle, RefreshCw, Zap, Webhook, } from "lucide-react"; import { cn } from "@/lib/utils"; import { useAdminRevenueStore } from "@/store/adminRevenueStore"; import { usePolling } from "@/hooks/usePolling"; import StatCard from "@/components/admin/StatCard"; import StatusDot from "@/components/admin/StatusDot"; import AlertBadge from "@/components/admin/AlertBadge"; import DataTable, { type ColumnDef } from "@/components/admin/DataTable"; import type { StripeStatus } from "@/store/adminRevenue… 证据：`gateway/frontend/src/app/admin/revenue/page.tsx`
- **Page**（source_file）：import { useEffect, useState, useCallback } from "react"; import { Calendar, Clock, AlertTriangle, RefreshCw, Plus, AlertCircle, CheckCircle2, Bell, FileText, } from "lucide-react"; import { cn } from "@/lib/utils"; import { getEvents, createEvent, type AdminEvent, type EventType, } from "@/services/api/adminService"; ⋮---- function formatEventDate iso: string : string ⋮---- function formatShortDate iso: string : string ⋮---- function isToday iso: string : boolean ⋮---- function isPast iso: string : boolean ⋮---- function EventTypeBadge ⋮---- className= ⋮---- const handleSubmit = async e: React.FormEvent = ⋮---- onChange= 证据：`gateway/frontend/src/app/admin/scheduler/page.tsx`
- **Page**（source_file）：import { useCallback, useEffect, useState } from "react"; import { usePolling } from "@/hooks/usePolling"; import StatusDot from "@/components/admin/StatusDot"; import { getSystemHealth, getSystemErrors, runSystemSelfTest, } from "@/services/api/adminService"; import type { SystemHealthReport, SystemErrorEntry, SelfTestResponse, SelfTestResult, SystemStatus, } from "@/services/api/adminService"; import { cn } from "@/lib/utils"; ⋮---- function formatUptime seconds: number : string ⋮---- function formatTimestamp iso: string : string ⋮---- function statusColor s: string : "green" "yellow" "red" "gray" ⋮---- function overallBannerClass status: SystemStatus : string ⋮---- function SkeletonBlock… 证据：`gateway/frontend/src/app/admin/system/page.tsx`
- **Page**（source_file）：import { useEffect, useState, useCallback } from "react"; import { Play, RefreshCw, CheckCircle2, XCircle, Clock, ChevronDown, ChevronUp, AlertCircle, Activity, Zap, } from "lucide-react"; import { cn } from "@/lib/utils"; import { getWorkflowTests, runWorkflowTest, runAllWorkflowTests, getTestHistory, type WorkflowTest, type RunTestResponse, type TestHistoryEntry, type TestResult, type TestStep, } from "@/services/api/adminService"; ⋮---- function formatTimestamp iso: string : string ⋮---- function formatMs ms: number : string ⋮---- className= ⋮---- onClick= ⋮---- <Play className= 证据：`gateway/frontend/src/app/admin/workflows/page.tsx`
- **Page**（source_file）：import { use, useState, useEffect, useCallback } from "react"; import Link from "next/link"; import { Store, Zap, CheckCircle, MessageCircle, ArrowLeft, User, Code2, QrCode, Copy, Check, RefreshCw, } from "lucide-react"; import { Button } from "@/components/ui/button"; import { Card } from "@/components/ui/card"; import { get } from "@/services/api/client"; ⋮---- interface PublicAgentProfile { handle: string; name: string; industry: string; description: string; agent type: string; avatar url: string; primary color: string; greeting: string; created at: string; capabilities: string ; card url: string; chat url: string; qr url: string; widget embed code: string; } ⋮---- interface PageProps {… 证据：`gateway/frontend/src/app/agents/[handle]/page.tsx`
- **Layout**（source_file）：import type { Metadata } from "next"; ⋮---- export default function AnalyticsLayout { children, }: { children: React.ReactNode; } 证据：`gateway/frontend/src/app/analytics/layout.tsx`
- **Page**（source_file）：import { useEffect, useState, useCallback, Component, type ErrorInfo, type ReactNode } from "react"; import Link from "next/link"; import { useRouter } from "next/navigation"; import { BarChart3, MessageSquare, Zap, DollarSign, AlertCircle, RefreshCw, Clock, TrendingUp, } from "lucide-react"; import { cn } from "@/lib/utils"; import { useAgentStore } from "@/store/agentStore"; import { ProtectedRoute } from "@/components/auth/ProtectedRoute"; import { AppShell } from "@/components/layout/AppShell"; import { PageHeader } from "@/components/layout/PageHeader"; import { getDashboardAnalytics, getConversations, getSkillUsage, getRevenue, type DashboardAnalytics, type ConversationEntry, type Ski… 证据：`gateway/frontend/src/app/analytics/page.tsx`
- **Page**（source_file）：import { Suspense, useEffect, useState } from "react"; import { useRouter, useSearchParams } from "next/navigation"; import { useAuthStore } from "@/store/authStore"; import { setAccessToken } from "@/services/api/client"; ⋮---- function OAuthCallbackHandler ⋮---- export default function OAuthCallbackPage 证据：`gateway/frontend/src/app/auth/callback/page.tsx`
- **Page**（source_file）：import { useEffect, Suspense } from 'react'; import { useRouter, useSearchParams } from 'next/navigation'; ⋮---- function HandoffInner ⋮---- // Set cookie so Next.js middleware auth guard recognises the session ⋮---- export default function HandoffPage 证据：`gateway/frontend/src/app/auth/handoff/page.tsx`
- **Page**（source_file）：import type { Metadata } from "next"; import Link from "next/link"; import Image from "next/image"; import { ArrowRight, Calendar, Shield, CheckCircle, Clock, TrendingUp, Car, Wrench, Phone, Bell, MessageSquare, FileText, } from "lucide-react"; ⋮---- function UseCaseCard { icon, title, description, }: { icon: React.ReactNode; title: string; description: string; } 证据：`gateway/frontend/src/app/automotive/page.tsx`
- **Layout**（source_file）：import type { Metadata } from "next"; ⋮---- export default function BillingLayout { children, }: { children: React.ReactNode; } 证据：`gateway/frontend/src/app/billing/layout.tsx`
- **Page**（source_file）：import { Suspense, useEffect, useState, useCallback } from "react"; import { useRouter, useSearchParams } from "next/navigation"; import { CreditCard, Zap, TrendingUp, AlertCircle, CheckCircle, ChevronUp, } from "lucide-react"; import { useAgentStore } from "@/store/agentStore"; import { getSkillUsage, getSkillUsageHistory, subscribeToPlan, createCheckoutSession, getSubscriptionStatus, type SkillUsageSummary, type SubscriptionStatusResponse, } from "@/services/api/paymentService"; import { ProtectedRoute } from "@/components/auth/ProtectedRoute"; import { AppShell } from "@/components/layout/AppShell"; import { PageHeader } from "@/components/layout/PageHeader"; import { Button } from "@/co… 证据：`gateway/frontend/src/app/billing/page.tsx`
- **Page**（source_file）：import { use, useState, useEffect, useRef, useCallback } from "react"; import Head from "next/head"; import Link from "next/link"; import { ArrowLeft, Send, RefreshCw, User, Store, Zap, AlertCircle, } from "lucide-react"; import { Button } from "@/components/ui/button"; ⋮---- interface AgentProfile { handle: string; name: string; industry: string; description: string; agent type: string; avatar url: string; primary color: string; greeting: string; } ⋮---- interface ChatMessage { id: string; role: "user" "assistant"; content: string; timestamp: Date; } ⋮---- interface PageProps { params: Promise ; } ⋮---- function generateId : string ⋮---- function formatHandle handle: string : string ⋮----… 证据：`gateway/frontend/src/app/chat/[handle]/page.tsx`
- **Layout**（source_file）：export default function ChatLayout { children, }: { children: React.ReactNode; } 证据：`gateway/frontend/src/app/chat/layout.tsx`
- **Page**（source_file）：import { useEffect } from "react"; import { useRouter } from "next/navigation"; ⋮---- export default function ChatPage 证据：`gateway/frontend/src/app/chat/page.tsx`
- **Layout**（source_file）：import type { Metadata } from "next"; ⋮---- export default function ClaimLayout { children, }: { children: React.ReactNode; } 证据：`gateway/frontend/src/app/claim/layout.tsx`
- **Page**（source_file）：import { useState, useEffect, useRef, useCallback } from "react"; import { useRouter } from "next/navigation"; import Image from "next/image"; import { ChevronRight, ChevronLeft, AlertCircle } from "lucide-react"; import { useAgentStore } from "@/store/agentStore"; import { Button } from "@/components/ui/button"; import { Input } from "@/components/ui/input"; import { Card } from "@/components/ui/card"; import { ProtectedRoute } from "@/components/auth/ProtectedRoute"; import { OnboardingProgress } from "@/components/onboarding/OnboardingProgress"; import { StepSector, type SectorItem } from "@/components/onboarding/StepSector"; import { StepTemplate, type TemplateItem } from "@/components/… 证据：`gateway/frontend/src/app/claim/page.tsx`
- **Page**（source_file）：import { useState, useEffect, useCallback } from 'react'; ⋮---- type RevenuePeriod = 'mrr' 'today' 'week' 'month'; ⋮---- interface NpmDownloads { downloads: number; package: string; } ⋮---- interface Product { name: string; slug: string; version: string; status: 'live' 'beta' 'down'; downloads: number; } ⋮---- interface Activity { id: string; avatar: string; text: string; time: string; } ⋮---- function Sparkline 证据：`gateway/frontend/src/app/dashboard/ceo/page.tsx`
- **Layout**（source_file）：import type { Metadata } from "next"; ⋮---- export default function DashboardLayout { children, }: { children: React.ReactNode; } 证据：`gateway/frontend/src/app/dashboard/layout.tsx`
- 其余 20 条证据见 `AI_CONTEXT_PACK.json` 或 `EVIDENCE_INDEX.json`。

## 宿主 AI 必须遵守的规则

- **把本资产当作开工前上下文，而不是运行环境。**：AI Context Pack 只包含证据化项目理解，不包含目标项目的可执行状态。 证据：`README.md`, `docs/architecture.md`, `gateway/frontend/README.md`
- **回答用户时区分可预览内容与必须安装后才能验证的内容。**：安装前体验的消费者价值来自降低误装和误判，而不是伪装成真实运行。 证据：`README.md`, `docs/architecture.md`, `gateway/frontend/README.md`

## 用户开工前应该回答的问题

- 你准备在哪个宿主 AI 或本地环境中使用它？
- 你只是想先体验工作流，还是准备真实安装？
- 你最在意的是安装成本、输出质量、还是和现有规则的冲突？

## 验收标准

- 所有能力声明都能回指到 evidence_refs 中的文件路径。
- AI_CONTEXT_PACK.md 没有把预览包装成真实运行。
- 用户能在 3 分钟内看懂适合谁、能做什么、如何开始和风险边界。

---

## Doramagic Context Augmentation

下面内容用于强化 Repomix/AI Context Pack 主体。Human Manual 只提供阅读骨架；踩坑日志会被转成宿主 AI 必须遵守的工作约束。

## Human Manual 骨架

使用规则：这里只是项目阅读路线和显著性信号，不是事实权威。具体事实仍必须回到 repo evidence / Claim Graph。

宿主 AI 硬性规则：
- 不得把页标题、章节顺序、摘要或 importance 当作项目事实证据。
- 解释 Human Manual 骨架时，必须明确说它只是阅读路线/显著性信号。
- 能力、安装、兼容性、运行状态和风险判断必须引用 repo evidence、source path 或 Claim Graph。

- **项目介绍**：importance `high`
  - source_paths: README.md, ARCHITECTURE.md, docs/innovations.md
- **快速开始**：importance `high`
  - source_paths: docs/quickstart.md, docker-compose.yml, config/agent.example.yaml, gateway/.env.example
- **系统架构**：importance `high`
  - source_paths: ARCHITECTURE.md, gateway/isg_agent/app.py, gateway/isg_agent/api/app.py, gateway/frontend/src/app/layout.tsx
- **数据流与状态管理**：importance `medium`
  - source_paths: gateway/isg_agent/db/schema.py, gateway/isg_agent/db/queries.py, gateway/isg_agent/db/engine.py, gateway/frontend/src/store/sessionStore.ts, gateway/frontend/src/store/chatStore.ts
- **七大创新特性**：importance `high`
  - source_paths: docs/innovations.md, config/constitution.example.yaml, gateway/isg_agent/schemas/audit.py
- **安全模型**：importance `high`
  - source_paths: docs/security-model.md, SECURITY.md, gateway/isg_agent/api/deps.py, gateway/isg_agent/auth/email_verification.py, gateway/isg_agent/config.py
- **代理宪法配置指南**：importance `high`
  - source_paths: docs/constitution-guide.md, config/constitution.example.yaml, gateway/isg_agent/api/routes/console_governance.py, gateway/isg_agent/schemas/skills.py
- **后端API系统**：importance `high`
  - source_paths: gateway/isg_agent/api/routes/agents.py, gateway/isg_agent/api/routes/auth.py, gateway/isg_agent/api/routes/templates.py, gateway/isg_agent/api/routes/tasks.py, gateway/isg_agent/api/routes/integrations.py

## Repo Inspection Evidence / 源码检查证据

- repo_clone_verified: true
- repo_inspection_verified: true
- repo_commit: `4fc48596d9895fc7609c4a9b4913996c2a5f626d`
- inspected_files: `pyproject.toml`, `package.json`, `README.md`, `docker-compose.yml`, `docs/innovations.md`, `docs/architecture.md`, `docs/security-model.md`, `docs/comparison-vs-openclaw.md`, `docs/api-reference.md`, `docs/quickstart.md`, `docs/constitution-guide.md`, `docs/skill-development.md`, `docs/deployment.md`

宿主 AI 硬性规则：
- 没有 repo_clone_verified=true 时，不得声称已经读过源码。
- 没有 repo_inspection_verified=true 时，不得把 README/docs/package 文件判断写成事实。
- 没有 quick_start_verified=true 时，不得声称 Quick Start 已跑通。

## Doramagic Pitfall Constraints / 踩坑约束

这些规则来自 Doramagic 发现、验证或编译过程中的项目专属坑点。宿主 AI 必须把它们当作工作约束，而不是普通说明文字。

### Constraint 1: 仓库名和安装名不一致

- Trigger: 仓库名 `dingdawg-agent-1` 与安装入口 `dingdawg-support-agent` 不完全一致。
- Host AI rule: 在 npm/PyPI/GitHub 上确认包名映射和官方 README 说明。
- Why it matters: 用户照着仓库名搜索包或照着包名找仓库时容易走错入口。
- Evidence: identity.distribution | mcp_registry:io.github.dingdawg/dingdawg-support-agent:2.0.6 | https://registry.modelcontextprotocol.io/v0.1/servers/io.github.dingdawg%2Fdingdawg-support-agent/versions/2.0.6 | repo=dingdawg-agent-1; install=dingdawg-support-agent
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 2: 能力判断依赖假设

- Trigger: README/documentation is current enough for a first validation pass.
- Host AI rule: 将假设转成下游验证清单。
- Why it matters: 假设不成立时，用户拿不到承诺的能力。
- Evidence: capability.assumptions | mcp_registry:io.github.dingdawg/dingdawg-support-agent:2.0.6 | https://registry.modelcontextprotocol.io/v0.1/servers/io.github.dingdawg%2Fdingdawg-support-agent/versions/2.0.6 | README/documentation is current enough for a first validation pass.
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 3: 维护活跃度未知

- Trigger: 未记录 last_activity_observed。
- Host AI rule: 补 GitHub 最近 commit、release、issue/PR 响应信号。
- Why it matters: 新项目、停更项目和活跃项目会被混在一起，推荐信任度下降。
- Evidence: evidence.maintainer_signals | mcp_registry:io.github.dingdawg/dingdawg-support-agent:2.0.6 | https://registry.modelcontextprotocol.io/v0.1/servers/io.github.dingdawg%2Fdingdawg-support-agent/versions/2.0.6 | last_activity_observed missing
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 4: 下游验证发现风险项

- Trigger: no_demo
- Host AI rule: 进入安全/权限治理复核队列。
- Why it matters: 下游已经要求复核，不能在页面中弱化。
- Evidence: downstream_validation.risk_items | mcp_registry:io.github.dingdawg/dingdawg-support-agent:2.0.6 | https://registry.modelcontextprotocol.io/v0.1/servers/io.github.dingdawg%2Fdingdawg-support-agent/versions/2.0.6 | no_demo; severity=medium
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 5: 存在评分风险

- Trigger: no_demo
- Host AI rule: 把风险写入边界卡，并确认是否需要人工复核。
- Why it matters: 风险会影响是否适合普通用户安装。
- Evidence: risks.scoring_risks | mcp_registry:io.github.dingdawg/dingdawg-support-agent:2.0.6 | https://registry.modelcontextprotocol.io/v0.1/servers/io.github.dingdawg%2Fdingdawg-support-agent/versions/2.0.6 | no_demo; severity=medium
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 6: issue/PR 响应质量未知

- Trigger: issue_or_pr_quality=unknown。
- Host AI rule: 抽样最近 issue/PR，判断是否长期无人处理。
- Why it matters: 用户无法判断遇到问题后是否有人维护。
- Evidence: evidence.maintainer_signals | mcp_registry:io.github.dingdawg/dingdawg-support-agent:2.0.6 | https://registry.modelcontextprotocol.io/v0.1/servers/io.github.dingdawg%2Fdingdawg-support-agent/versions/2.0.6 | issue_or_pr_quality=unknown
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。

### Constraint 7: 发布节奏不明确

- Trigger: release_recency=unknown。
- Host AI rule: 确认最近 release/tag 和 README 安装命令是否一致。
- Why it matters: 安装命令和文档可能落后于代码，用户踩坑概率升高。
- Evidence: evidence.maintainer_signals | mcp_registry:io.github.dingdawg/dingdawg-support-agent:2.0.6 | https://registry.modelcontextprotocol.io/v0.1/servers/io.github.dingdawg%2Fdingdawg-support-agent/versions/2.0.6 | release_recency=unknown
- Hard boundary: 不要把这个坑点包装成已解决、已验证或可忽略，除非后续验证证据明确证明它已经关闭。