git-repo-brief
MCP server that provides structured briefs for public GitHub repositories, including metadata, key files, release notes, and activity snapshots.
README
git-repo-brief
An MCP (Model Context Protocol) server that generates structured briefs for public GitHub repositories. Get comprehensive repository information including overview, key files, release notes, and activity snapshots.
Features
- repo_overview: Get repository metadata (name, description, stars, forks, language, topics, license, timestamps)
- extract_key_files: Fetch contents of key files (README.md, LICENSE, package.json, pyproject.toml)
- release_notes: Get recent release notes with tag names, descriptions, and dates
- activity_snapshot: Get activity metrics (last commit, open issues/PRs, contributors count)
Installation
npm install
npm run build
Usage
As MCP Server (stdio mode)
# Run directly
npm start
# Or after building
node dist/cli.js
As HTTP Server
npm start -- --http
# or
HTTP_PORT=3001 node dist/cli.js --http
MCP Client Configuration
Add to your MCP client configuration (e.g., Claude Desktop):
{
"mcpServers": {
"git-repo-brief": {
"command": "node",
"args": ["/path/to/git-repo-brief/dist/cli.js"]
}
}
}
Tools
repo_overview
Get a comprehensive overview of a GitHub repository.
Input:
{
"repo_url": "https://github.com/owner/repo"
}
Output:
{
"ok": true,
"data": {
"name": "repo",
"full_name": "owner/repo",
"description": "Repository description",
"stars": 1234,
"forks": 567,
"language": "TypeScript",
"topics": ["mcp", "github"],
"license": "MIT",
"created_at": "2023-01-01T00:00:00Z",
"updated_at": "2024-01-01T00:00:00Z",
"default_branch": "main",
"homepage": "https://example.com",
"owner": {
"login": "owner",
"avatar_url": "https://github.com/owner.png",
"type": "User"
}
},
"meta": {
"source": "https://api.github.com/repos/owner/repo",
"retrieved_at": "2024-01-15T10:00:00Z",
"pagination": { "next_cursor": null },
"warnings": []
}
}
extract_key_files
Fetch contents of key files from a repository.
Input:
{
"repo_url": "https://github.com/owner/repo",
"paths": ["README.md", "package.json"]
}
Output:
{
"ok": true,
"data": {
"files": [
{
"path": "README.md",
"content": "# Project\n\nDescription...",
"size": 1234,
"encoding": "utf-8"
},
{
"path": "package.json",
"content": "{\"name\": \"project\"}",
"size": 456,
"encoding": "utf-8"
}
],
"default_branch": "main"
},
"meta": {
"retrieved_at": "2024-01-15T10:00:00Z",
"warnings": []
}
}
release_notes
Get recent release notes from a repository.
Input:
{
"repo_url": "https://github.com/owner/repo",
"limit": 5
}
Output:
{
"ok": true,
"data": {
"releases": [
{
"tag_name": "v1.0.0",
"name": "Version 1.0.0",
"body": "## Changes\n- Feature A\n- Bug fix B",
"published_at": "2024-01-01T00:00:00Z",
"prerelease": false,
"draft": false,
"html_url": "https://github.com/owner/repo/releases/tag/v1.0.0"
}
],
"total_count": 1
},
"meta": {
"retrieved_at": "2024-01-15T10:00:00Z",
"pagination": { "next_cursor": null },
"warnings": []
}
}
activity_snapshot
Get a snapshot of recent repository activity.
Input:
{
"repo_url": "https://github.com/owner/repo"
}
Output:
{
"ok": true,
"data": {
"last_commit_date": "2024-01-14T15:30:00Z",
"last_commit_message": "fix: resolve parsing issue",
"open_issues_count": 25,
"open_prs_count": 8,
"contributors_count": 42,
"watchers_count": 100,
"has_wiki": true,
"has_discussions": true
},
"meta": {
"retrieved_at": "2024-01-15T10:00:00Z",
"warnings": []
}
}
HTTP API Endpoints
When running in HTTP mode:
GET /health- Health checkGET /tools- List available toolsPOST /tools/repo_overview- Get repository overviewPOST /tools/extract_key_files- Extract key filesPOST /tools/release_notes- Get release notesPOST /tools/activity_snapshot- Get activity snapshot
Example HTTP Request
curl -X POST http://localhost:3000/tools/repo_overview \
-H "Content-Type: application/json" \
-d '{"repo_url": "https://github.com/microsoft/vscode"}'
Configuration
Environment Variables
| Variable | Default | Description |
|---|---|---|
GITHUB_TOKEN |
- | GitHub personal access token (optional, increases rate limit from 60 to 5000 req/hour) |
REQUEST_DELAY_MS |
100 | Delay between requests for rate limiting |
REQUEST_TIMEOUT_MS |
30000 | Request timeout in milliseconds |
TRANSPORT_MODE |
stdio | Transport mode: stdio or http |
HTTP_PORT |
3000 | HTTP server port (when using HTTP transport) |
Rate Limiting
The server implements polite rate limiting:
- Default delay of 100ms between requests
- Handles GitHub API rate limit responses gracefully
- Optional GitHub token support for higher rate limits
Response Format
All responses follow the standard envelope format:
Success Response
{
"ok": true,
"data": { ... },
"meta": {
"source": "optional API URL",
"retrieved_at": "ISO-8601 timestamp",
"pagination": { "next_cursor": null },
"warnings": []
}
}
Error Response
{
"ok": false,
"error": {
"code": "INVALID_INPUT | UPSTREAM_ERROR | RATE_LIMITED | TIMEOUT | PARSE_ERROR | INTERNAL_ERROR",
"message": "Human readable message",
"details": { ... }
},
"meta": {
"retrieved_at": "ISO-8601 timestamp"
}
}
Development
# Install dependencies
npm install
# Run in development mode
npm run dev
# Run tests
npm test
# Run tests with coverage
npm run test:coverage
# Type check
npm run typecheck
# Build
npm run build
Testing
# Run all tests
npm test
# Run tests in watch mode
npm run test:watch
# Run with coverage
npm run test:coverage
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。