Context7-ChatGPT-Bridge

This MCP server acts as a bridge between ChatGPT and Context7's current documentation database. It exposes search and fetch capabilities so ChatGPT can discover relevant programming libraries, then retrieve up-to-date documentation through Context7. The bridge translates ChatGPT’s tools into Context7 API calls (resolve-library-id and get-library-docs) and formats the results for ChatGPT to present.

To use it, run the bridge locally or in your environment (it will expose an HTTP/SSE endpoint that ChatGPT can connect to via ngrok). Once running, you can use the two primary tools:

• search: Look up libraries or Context7 IDs by name or direct IDs. For example searching for "React" or "/reactjs/react.dev" returns library IDs and metadata.
• fetch: Retrieve comprehensive documentation for a chosen library. You can request a basic doc (library_id), or deeper content using tokens, topics, or a combination like library_id|topic:authentication|tokens:12000 to tailor the depth and scope.

ChatGPT communicates with the bridge, which then calls Context7’s documentation database and returns formatted results suitable for follow-up questions, code examples, or API references. The system also handles common workflow requirements, such as performing a search before fetch to ensure valid IDs are used.

Prerequisites:

• Node.js is not required for this bridge (the bridge is Python-based), but ensure Python >= 3.8 is installed.
• Python 3.8+ and pip are required.
• Optional: ngrok if you want a public URL for ChatGPT access (the bridge can run with or without ngrok).

Installation steps:

1. Clone the repository or download the README/example project files to your environment.
2. Create and activate a Python virtual environment (recommended):
   • python3 -m venv venv
   • source venv/bin/activate # macOS/Linux
   • venv\Scripts\activate # Windows
3. Install dependencies:
   • pip install -r requirements.txt
4. Run the bridge:
   • python context7_bridge.py The bridge will start on the default port (8000) and, if ngrok is available, will automatically expose a public URL for ChatGPT integration.
5. Optional: If you prefer manual ngrok management:
   • Run the bridge with no-ngrok: python context7_bridge.py --no-ngrok
   • In another terminal: ngrok http 8000
6. Add the ngrok URL to ChatGPT as an MCP connector (for example: https://abc123.ngrok-free.app/sse).

Tips:

• Always perform a search first to obtain valid library IDs before using fetch. If you encounter "Unknown document ID", retry with a search or use a direct Context7 library ID starting with '/'.
• You can customize the depth of fetched documentation using the tokens parameter and topic filters (e.g., topic: hooks, authentication, installation, api, examples).
• The bridge expects JSON-RPC style requests from ChatGPT via the SSE endpoint; ensure your ChatGPT connector is configured to use the /sse route.
• If you run into connectivity issues, verify that Python, pip, and network access to Context7 are available. You can enable debug logging by starting the bridge with: LOG_LEVEL=DEBUG python context7_bridge.py
• If ngrok is not desired, you can expose a local 8000 port behind your own reverse proxy or VPN; the bridge will operate normally without ngrok.