菜单
Download documentation from GitHub repos or URLs. Supports manifest-based batch downloads with configurable exclude patterns.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Local Microsoft Agent Framework documentation reference. Use when asked about Microsoft Agent Framework, building AI agents in .NET or Python, MCP servers/clients, durable agents, agent tools, Teams/WebChat adapters, or agent-to-agent communication.
Generate documentation skills for AI agents from downloaded docs. Use after /build-my-context7 to create skills.
Local LangChain AI documentation reference. Use when asked about LangChain, LangGraph, agents, chains, prompts, memory, tools, retrieval, RAG, vector stores, document loaders, or building LLM applications.
Local Next.js documentation reference (canary branch). Use when asked about Next.js features, App Router, Server Components, routing, data fetching, rendering, caching, styling, optimizations, configuration, or Next.js APIs.
Local Prisma documentation reference. Use when asked about Prisma ORM, Prisma Client, Prisma Schema, migrations, database queries, Prisma Accelerate, or Prisma Postgres.
Local TanStack Form documentation reference. Use when asked about TanStack Form features, form validation, form state management, field arrays, React/Vue/Solid/Angular/Svelte form integration, or form submission handling.
| name | download-docs |
| description | Download documentation from GitHub repos or URLs. Supports manifest-based batch downloads with configurable exclude patterns. |
| user-invocable | false |
Download documentation files from various sources using manifest configurations.
Download from a specific manifest:
.claude/skills/download-docs/scripts/download.sh {manifest-name}
Download from all manifests:
.claude/skills/download-docs/scripts/download.sh
Output location: output/{manifest-name}/
Manifests are JSON files in scripts/manifests/. Use $schema for IDE validation. Two source types are supported:
Clones repository and extracts markdown files:
{
"$schema": "./manifest.schema.json",
"_source": {
"type": "github",
"repo": "owner/repo-name",
"branch": "main",
"path": "docs",
"extensions": [".md", ".mdx"],
"exclude": ["**/internal/**"],
"noDefaultExcludes": false
}
}
| Field | Required | Description |
|---|---|---|
repo | Yes | GitHub repository (owner/name) |
branch | No | Branch to clone (default: "main") |
path | No | Directory path within repo (default: ".") |
extensions | No | File extensions to include (default: [".md", ".mdx"]) |
exclude | No | Additional glob patterns to exclude |
noDefaultExcludes | No | Set true to disable default excludes |
Default exclude patterns:
Downloads specific files from URLs:
{
"$schema": "./manifest.schema.json",
"_source": { "type": "url" },
"files": {
"category-name": {
"doc-name": "https://example.com/path/to/doc.md"
}
}
}
Downloads HTML and converts to markdown (requires pandoc):
{
"$schema": "./manifest.schema.json",
"_source": {
"type": "url",
"convert": "html"
},
"files": {
"category-name": {
"doc-name": "https://example.com/docs/page.html"
}
}
}
Note: HTML conversion only works for static HTML. SPA sites (React, Angular, etc.) will produce empty output.
Add _disabled to skip processing:
{
"$schema": "./manifest.schema.json",
"_source": {
"type": "url",
"_disabled": true,
"_reason": "SPA site - requires Playwright handler"
},
"files": { ... }
}
jq - JSON parsinggit - GitHub source cloningcurl - URL downloadspandoc - HTML conversion (optional)Create a JSON file in scripts/manifests/ following the format above. The manifest filename (without .json) becomes the output directory name.