github-pr-context-mcp 0.2.8

GitHub PR Review Context MCP Server

GitHub PR Review Context MCP

Production-grade context layer for AI code review, grounded in your repository's real pull request history.

---

Overview

GitHub PR Review Context MCP gives AI assistants institutional review memory.

Instead of generic feedback, reviews are informed by historical reviewer comments, recurring quality patterns, and repository-specific standards from your own PR history.

Core Value

• Improves review consistency across teams and repositories.
• Reduces repeated reviewer feedback on known issues.
• Integrates with any MCP-compatible client and multiple LLM providers.

---

Key Capabilities

Capability	What It Delivers
Historical review retrieval	Semantic search across prior PR comments and review summaries
Context-aware AI review	Feedback grounded in repository-specific review behavior
Grounded code generation	Generate new code based on past commits, comments, and style
Team rules generation	Auto-generate .cursorrules / CLAUDE.md from repo history
Smart repository readiness	Auto-detect indexed state and index on demand
Flexible storage modes	Permanent (disk) and temporary (in-memory) indexing options
Portable inference layer	Switch LLM providers using environment configuration only

---

🚀 Quick Start

🚀 Zero-Setup (uvx / pipx / npx)

The fastest way to use the server. No cloning required. Just run one of these commands directly in your terminal or use them in your IDE's MCP settings:

[!TIP] Don't clone this repo to get AI rules! Once installed, run generate_repo_rules inside YOUR project to automatically create .cursorrules or CLAUDE.md tailored to your own team's PR history.

Using uvx (Recommended for speed):

uvx github-pr-context-mcp

Using pipx (Recommended for stability):

pipx run github-pr-context-mcp
# Or install permanently:
pipx install github-pr-context-mcp

Using npx (Smithery bridge):

npx -y @smithery/cli run github-pr-context-mcp

---

⚠️ Manual Installation (Git Clone / Advanced)

[!WARNING] Running from a git clone is only recommended for developers contributing to this project. For general use, please use the pipx method above.

If you have cloned the repository for development:

1. Create a virtual environment: python -m venv .venv
2. Activate it and install: pip install -e .
3. Run automatic setup: python scripts/install_clients.py

For full configuration (Cursor, Claude Desktop), see the Quick Start Guide.

---

[!IMPORTANT] 🚀 USE THE OFFICIAL PACKAGE: This project is now on PyPI. To ensure seamless updates and zero configuration friction, do NOT git clone.

Recommended Install:

pipx install github-pr-context-mcp

Or run instantly with: uvx github-pr-context-mcp

---

🛠️ Usage Modes: Solo vs. Team

This MCP server is built to scale from a single machine to an entire engineering organization.

👤 Solo Developer (Local Mode)

Best for: Privacy, local-first control, and zero hosting costs.

• How it works: Run via uvx, pipx, or a local git clone.
• Storage: ChromaDB stays on your local machine.
• Security: Your GitHub Token and LLM keys never leave your device.
• Setup: See Quick Start.

🤝 Team Collaboration (Hosted Mode — UPCOMING)

Best for: Scaling team-wide PR standards and centralized infra.

• How it works: One deployment on Render (Coming Soon) shared by the whole team.
• Isolation: Strict Gmail-based namespace isolation (driven by SQLite). User A's indexed data is mathematically invisible to User B.
• Economics: Pooled LLM credits and a single shared indexing server.
• Setup: See Deployment Guide.

🌟 Zero-Friction Setup (Upcoming)

If your team has hosted this MCP on Render, you do NOT need to git clone or install anything. Just drop this snippet into your IDE:

"github-pr-context": {
  "type": "sse",
  "url": "https://YOUR-RENDER-URL.onrender.com/mcp",
  "headers": {
    "Authorization": "Bearer YOUR_TOKEN"
  }
}

That's it. If your IDE supports native MCP SSE connections, you are immediately connected to the secure Render deployment.

---

🧰 Core Tools Reference

The server exposes 12 core tools for IDE agents and developers. For a deep dive on when to use each, see the Tool Strategy Guide.

Tool	Action
ensure_repo_ready	Index a repo and ensure it's ready for queries
generate_repo_rules	Synthesize .cursorrules / CLAUDE.md from PR history
generate_code_from_history	Write code grounded in past commits & team style
review_code_with_history	Perform AI review grounded in team review memory
get_team_review_patterns	Summarize recurring team standards (e.g. "no magic numbers")
semantic_search_reviews	Search past PR comments by meaning, not just keywords
set_active_repo	Switch between multiple indexed repositories
list_indexed_repos	View all repos currently in local/temporary storage
delete_repo_index	Free up disk space by clearing repository indices
get_index_stats	Verify if a repo index is complete (doc count)
update_settings	Update tokens/LLM keys (Hosted mode only)
get_usage_stats	View adoption metrics and unique user counts

---

📖 Documentation

Detailed guides for deep dives and specific configurations:

• 🛠️ Quick Start & Usage — Setup and basic commands.
• ⚙️ LLM Configuration — Switching between OpenAI, Anthropic, Gemini, and Cerebras.
• 🧩 Tool Strategy & Selection Guide — When to use which tool (for humans and agents).
• 🏗️ Architecture & Pipeline — How the RAG engine and indexing work.
• 🔌 Integrations — Connecting to Cursor, Claude Desktop, and more.

---

🛠️ Troubleshooting

• "command not found": Use absolute paths in your configuration. Run github-pr-context-mcp config to get your exact path.
• "PermissionError: [WinError 32]": The binary is locked by a running process. Close Claude/Cursor, run taskkill /F /IM github-pr-context-mcp.exe, then retry the upgrade.
• Rate Limit Errors: Ensure your GITHUB_TOKEN is valid and has repo scope.

---

📣 Community & Feedback

We want to hear from you — whether you are a solo developer or a team at a large company!

• Feedback: Please open an issue or start a discussion if you have ideas or encounter bugs.
• Star ⭐: If this tool saves you time, give it a star! It helps others find the project.
• Corporate: Is your team using this? Join our "Adopters" list by opening a PR to add your team's name.

---

⚖️ License

MIT