mcp -exploration

This MCP server acts as a bridge between a local OpenAPI/Swagger-based API and an MCP client, enabling dynamic tool registration and execution via the Connect API. It leverages the OpenAPI MCP tooling to register available operations as tools, allowing the client to discover and invoke them through the MCP interface. The README indicates you run a server with authentication and a Swagger specification, then run a separate client that exercises or consumes the registered tools. The included Swagger tooling snippet demonstrates how the server integrates Swagger-defined operations into tools that can be invoked by the client.

Prerequisites:

• Python 3.8+ installed on your system
• Git installed
• Access to a Swagger/OpenAPI spec (swagger.yaml)

1. Clone the repository (or set up your project directory): git clone https://github.com/<owner>/<repo>.git cd <repo>

2. Install Python dependencies (adjust if you use a virtual environment): python -m venv venv source venv/bin/activate # On Windows use: venv\Scripts\activate pip install -r requirements.txt

3. Prepare environment variables and Swagger file:

   • Create or point to your swagger.yaml file that describes your API.
   • Ensure you have a Connect API key for authentication.

4. Run the MCP server: Connect_API_KEY="<your key>" SWAGGER_FILE="swagger.yaml" make server

5. In another terminal, run the MCP client (as indicated in the README): make client

Notes:

• The exact module or script name for app.py may vary; ensure you’re invoking the server entry point correctly (the README references app.py with a main function).
• If you prefer a direct Python invocation instead of make, you can run: python app.py
• Adjust paths to swagger.yaml as needed for your environment.

Tips and common issues:

• Ensure SWAGGER_FILE points to a valid OpenAPI/Swagger spec compatible with the tooling used by the server.
• The server expects a valid Connect API key; missing or invalid keys will prevent initialization.
• If the client cannot discover tools, verify that the Swagger transform utilities (transform_swagger_to_operation_dict, map_operations_to_tools) are correctly importing from openapi_mcp and that the Swagger doc is accessible at the expected endpoints.
• If you modify app.py to change the questions or flow, re-run the server so the registered tools reflect the new behavior.
• Environment variables can be added or overridden per deployment; consider using a .env file or your process manager’s env config for production deployments.