OSSInsight MCP Server
A Model Context Protocol (MCP) server for querying GitHub repository analytics from OSSInsight, providing repo metadata, trending repos, curated collections, search, compare, and collection rankings without an API key.
README
OSSInsight MCP Server
A Model Context Protocol (MCP) server for querying GitHub repository analytics from OSSInsight (by PingCAP). Provides repo metadata, trending repos, curated collections, search, compare, and collection rankings — no API key required.
Features
ossinsight_repo— GitHub repo metadata: stars, forks, language, license, descriptionossinsight_trending— Trending repos by language and timeframeossinsight_collections— List all 102 curated collectionsossinsight_search— Search repos and collections by keywordossinsight_compare— Side-by-side comparison of two reposossinsight_ranking— Rank repos in a collection by stars, issues, or PRs
Installation
uv (recommended)
uvx ossinsight-mcp
pip
pip install ossinsight-mcp
ossinsight-mcp
Configuration
Claude Desktop
{
"mcpServers": {
"ossinsight-mcp": {
"command": "uvx",
"args": ["ossinsight-mcp"]
}
}
}
Cursor / VS Code
{
"mcpServers": {
"ossinsight-mcp": {
"command": "uvx",
"args": ["ossinsight-mcp"]
}
}
}
Tools
| Tool | Description | Key Params |
|---|---|---|
ossinsight_repo |
Repo metadata — stars, forks, language, license | owner, repo (required) |
ossinsight_trending |
Trending repos by language/timeframe | language, period (past_24h/past_week/past_month/past_3mo) |
ossinsight_collections |
List all curated collections | (none) |
ossinsight_search |
Search repos and collections | q (required), page, page_size, sort |
ossinsight_compare |
Compare two repos side-by-side | repo1, repo2 (required) |
ossinsight_ranking |
Rank repos in a collection | collection_id (required), metric, range |
Examples
# Get repo metadata
ossinsight_repo({"owner": "grafana", "repo": "grafana"})
# → stars: 76K, forks: 14K, language: TypeScript, license: AGPL-3.0
# What's trending in Rust this month
ossinsight_trending({"language": "rust", "period": "past_month"})
# List all collections
ossinsight_collections({})
# Compare two databases
ossinsight_compare({"repo1": "postgres/postgres", "repo2": "clickhouse/clickhouse"})
# Top repos in the AI/ML collection by stars this year
ossinsight_ranking({"collection_id": 12, "metric": "stars", "range": "year"})
Data Source
Uses the OSSInsight API (/api/mcp).
- Auth: None required
- Rate limit: 600 requests per hour per IP
- Data: GH Archive (all public GitHub events since 2011)
Development
git clone https://github.com/GeniusTechnoMystic/ossinsight-mcp.git
cd ossinsight-mcp
uv sync
uv run ossinsight-mcp
Related Projects
Other OSSInsight MCP servers in the ecosystem:
| Project | Language | Tools | Tests | Install | Differentiator |
|---|---|---|---|---|---|
| damonxue/mcp-ossinsight | Python | 5 | None | npx / Docker |
Developer analysis, natural language query, web scraping fallback |
| gztchan/ossinsight-mcp | Python | 4 | None | git clone | Stargazer/issue/PR creator demographics, generic request tool |
| This one (ours) | Python | 6 | 28 tests | uvx / pip install |
Collections, search, compare, ranking, PyPI-published |
Our differentiators
- Broadest open-source tool set — 6 tools covering repo metadata, trending, collections, search, compare, and ranking. The only open-source server with dedicated tools for collections, search, and compare.
- Test coverage — 28 tests with mocked HTTP (respx). No competitor has tests.
- PyPI publication — install with
uvx ossinsight-mcporpip install ossinsight-mcp. No npm, no Docker, no build step. - Ranking flexibility —
ossinsight_rankingsupports stars, issues, or PRs across day/week/month/year ranges.
When to use the alternatives
- damonxue/mcp-ossinsight — if you need developer contribution analysis, natural language queries, or web scraping for repos with <25 stars.
- gztchan/ossinsight-mcp — if you need stargazer/issue/PR creator demographics (countries, history) or a generic forward-compatible request tool.
Scope note
None of the open-source OSSInsight MCP servers fully wrap the OSSInsight Public API v1beta. A hosted 18-tool OSSInsight MCP server exists on mcpbundles.com covering issue creator, PR creator, and stargazer demographics (countries, history, orgs) — but it's not open source.
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。