OSS Intelligence MCP Server

An MCP server that provides AI-native open-source repository intelligence. Ask questions about any GitHub repository and get structured, visual answers with interactive UI panels — directly inside Cursor, Claude Desktop, VS Code, or any MCP-compatible client.

What It Does

Three tools answer two fundamental questions:

• "Should I use this?" — analyze_repo produces a comprehensive health report with an overall verdict (Safe / Caution / Risky), covering activity trends, bus factor, PR health, issue responsiveness, release cadence, corporate backing, and security score.
• "Should I contribute?" — should_i_contribute surfaces contributor experience signals: PR merge rates, review times, maintainer responsiveness, good first issues, and contributor retention.
• "Which one is better?" — compare_repos runs side-by-side analysis of 2–3 repos with per-metric winners and a recommendation.

Data comes from the GitHub REST API (primary) enriched by three free external APIs:

• deps.dev — Google's dependency intelligence (includes OpenSSF Scorecard)
• OpenSSF Scorecard — automated security assessment
• OSS Insight — pre-computed contributor affiliation data

All external APIs are optional — the server degrades gracefully to GitHub-only if any are unreachable.

Quick Start

docker run -d -p 9847:9847 -e GITHUB_TOKEN=ghp_your_token_here ghcr.io/jonzarecki/oss-intel-mcp

Then configure your MCP client to connect:

Cursor — add to ~/.cursor/mcp.json:

{
  "mcpServers": {
    "oss-intel": {
      "url": "http://localhost:9847/mcp"
    }
  }
}

Claude Desktop — add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "oss-intel": {
      "url": "http://localhost:9847/mcp"
    }
  }
}

Get a GitHub Token

Create a personal access token at github.com/settings/tokens. The only scope needed is public_repo (read-only access to public repos).

Tools

analyze_repo(owner, repo)

Full health report with overall verdict.

Example prompt: "Analyze the health of expressjs/express"

Returns: Verdict (Safe/Caution/Risky), score (0–100), 8 metric breakdowns, enrichment source indicators.

UI Panel: "Should I Use This?" verdict card with stats grid, breakdown chart, corporate backing, and security score.

should_i_contribute(owner, repo)

Contributor experience analysis.

Example prompt: "Should I contribute to facebook/react?"

Returns: PR merge rates, review times, maintainer responsiveness, good first issue count, contributor retention.

UI Panel: "Is It Worth Contributing?" dashboard with PR funnel, review gauge, and retention bar.

compare_repos(repos)

Side-by-side comparison of 2–3 repos.

Example prompt: "Compare express, fastify, and koa"

Returns: Per-metric comparison table with winners, overall recommendation.

Metrics

Metric	Method	Weight in Verdict
Activity Trend	3-month commit volume comparison	20%
Bus Factor	Gini coefficient of commit distribution	17%
PR Health	Merge rate + review time + merge time	17%
Security	OpenSSF Scorecard (17 checks)	17%
Issue Responsiveness	Response time + close rate	13%
Release Cadence	Interval regularity + frequency	8%
Corporate Backing	Org diversity + contribution share	8%

When security data is unavailable, remaining metrics re-normalize to 100%.

License

MIT