skills/agents-md-generator/SKILL.md
Generate comprehensive AGENTS.md documentation for backend projects with complete API specifications, business rules, data models, and data flows. Use when (1) Creating AGENTS.md from existing CLAUDE.md, (2) Documenting backend API modules with FastAPI routes, (3) Migrating documentation to AGENTS.md/CLAUDE.md symlink structure, (4) Adding complete API interface documentation to existing specs, (5) Creating module-level AGENTS.md for specific features (mcp, teamo_code, file_system, etc.)
npx skillsauth add realroc/skills agents-md-generatorInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Generate comprehensive AGENTS.md documentation for backend projects following the AGENTS.md/CLAUDE.md symlink pattern.
Ask the user to clarify:
Read all relevant source files:
clawdbot.py, router.py)Use Read tool to load files into context. For large codebases, use Grep to find specific patterns first.
Follow the template structure in references/agents_template.md:
Required Sections:
references/api_doc_template.md)references/data_flow_template.md)Content Priorities:
After creating AGENTS.md:
cp CLAUDE.md CLAUDE.md.backuprm CLAUDE.mdln -s AGENTS.md CLAUDE.mdls -la CLAUDE.md should show CLAUDE.md -> AGENTS.mdUse the Bash tool for these operations.
Structure Validation:
[Section](#section))Content Validation:
Symlink Validation:
# Verify symlink
ls -la CLAUDE.md
# Verify content match
diff CLAUDE.md AGENTS.md
# Expected: No output (files are identical via symlink)
Follow the project's Git commit standards:
git add AGENTS.md CLAUDE.md CLAUDE.md.backup
git commit -m "docs: 创建 <模块名称> AGENTS.md 并设置 CLAUDE.md 软链接
- 创建 AGENTS.md (完整的模块规范文档)
- 包含 <X> 个接口的完整文档
- 包含业务规范 (<业务规范名称>)
- 包含数据模型和外部集成说明
- 创建软链接 CLAUDE.md -> AGENTS.md
- 备份原 CLAUDE.md 为 CLAUDE.md.backup
Co-Authored-By: Claude Sonnet 4.5 <[email protected]>"
For each API interface, document:
Basic Info:
Request:
Response:
Business Logic (if complex):
See references/api_doc_template.md for detailed examples.
Use ASCII diagrams with clear stages:
用户操作
↓
API 调用 (接口名称)
├─ Step 1: 验证 (描述)
├─ Step 2: 数据库操作 (描述)
│ ├─ 查询 xxx
│ └─ 更新 yyy
├─ Step 3: 外部服务调用 (描述)
└─ Step 4: 返回响应
For complex flows, break into multiple sub-diagrams. See references/data_flow_template.md for examples.
Root AGENTS.md: Global specs, cross-module patterns, project overview Module AGENTS.md: Module-specific specs, API docs, detailed flows
Reference module docs from root:
请参考对应模块的文档获取详细的开发指南:
- **MCP 模块**: `wowchat/mcp/AGENTS.md`
- **Teamo Code 模块**: `wowchat/teamo_code/AGENTS.md`
Update Spec auto-deposition mechanism to use AGENTS.md:
每次修改 AGENTS.md 后,必须在回复尾部单独一行输出(前后空行):
\`\`\`
喵 ฅ՞•ﻌ•՞ฅ 已完成xxx相关的spec,更新/删除AGENTS.md文件中相关内容。
\`\`\`
For module-level updates:
喵 ฅ՞•ﻌ•՞ฅ 已完成xxx相关的spec,更新/删除<模块路径>/AGENTS.md文件中相关内容。
When migrating from CLAUDE.md to AGENTS.md, remove:
Preserve:
For projects with multiple modules, create hierarchy:
/AGENTS.md (root - global specs)
/CLAUDE.md -> AGENTS.md (symlink)
/wowchat/mcp/AGENTS.md (module specs)
/wowchat/mcp/CLAUDE.md -> AGENTS.md (symlink)
/wowchat/teamo_code/AGENTS.md (module specs)
/wowchat/teamo_code/CLAUDE.md -> AGENTS.md (symlink)
For each external service, document:
See MCP module example for Lighthouse, Feishu, Alipay integrations.
Common Mistakes:
ls -laValidation Checklist:
ls -la CLAUDE.md shows -> AGENTS.md)cat CLAUDE.md shows AGENTS.md contentreferences/agents_template.mdreferences/api_doc_template.mdreferences/data_flow_template.mdassets/example_mcp_agents.mddevelopment
Screen MongoDB conversation collections for script-driven abuse (prompt-injection templates, curl/empty user agents, probe-word floods, sessionless calls, multi-account IPs). Produces a two-tier triage report (confirmed abuse / suspicious) plus a multi-account IP list and a ban candidate CSV. Use when asked to find script callers, prompt-injection attempts, abnormal high-frequency users, accounts bypassing the web UI, or "who is using my AI as a cron job".
development
Audit or rewrite a prompt into a six-section issue spec (Goal / Constraints / Non-goals / Verification / Architecture notes / Existing context) before any code gets generated. Use when the user pastes a vague request and asks for implementation, or explicitly says they want to frame an issue properly. Triggers on: prompt spec, audit this prompt, check my prompt, what's missing in this prompt, frame this issue, rewrite as a prompt spec, convert to issue spec, make this an issue, issue framing.
testing
GitHire's six-step AI-native engineering method: frame the issue, sandbox, AI execute, AI review, architect decision, ship. Use when planning or executing real work with AI agents — issue framing, prompt writing, PR review gating, architect handoff — or anytime humans-frame-AI-execute-architects-verify applies. Triggers on: use githire, githire methodology, issue-first onboarding, ai-native workflow, frame this issue, prompt spec, architect review, first PR for a candidate, hire through real PRs.
development
Geolocate a batch of IPv4 addresses and produce a Markdown distribution table — Chinese IPs broken down by province (incl. HK/MO/TW), foreign IPs by country, with counts and percentages. Optionally exports CSV. Uses the free ip-api.com batch endpoint (no key, no signup, HTTP only, 15 batches × 100 IPs per minute). Use when the user pastes a list of IPs and asks for "IP 分布", "IP 归属地分布", "省份分布", "where are these IPs from", "geolocate these IPs", or wants an IP-region breakdown table.