README.md•3.06 kB
# OpenAPI MCP Server
[](https://smithery.ai/server/@rahgadda/openapi_mcp_server)
## Overview
- This project will install `MCP - Model Context Protocol Server`, that provides configured REST API's as context to LLM's.
- Using this we can enable LLMs to interact with RestAPI's and perform REST API call's using LLM prompts.
- Currently we support HTTP API Call's `GET/PUT/POST/PATCH`.
## Installation
- Install package
```bash
pip install openapi_mcp_server
```
- Create .env in a folder with minimum values for `OPENAPI_SPEC_PATH` & `API_BASE_URL`. Sample file available [here](https://raw.githubusercontent.com/rahgadda/openapi_mcp_server/refs/heads/main/.env)
- Test `openapi_mcp_server` server using `uv run openapi_mcp_server` from the above folder.
## Claud Desktop
- Configuration details for Claud Desktop
```json
{
"mcpServers": {
"openapi_mcp_server":{
"command": "uv",
"args": ["run","openapi_mcp_server"]
"env": {
"DEBUG":"1",
"API_BASE_URL":"https://petstore.swagger.io/v2",
"OPENAPI_SPEC_PATH":"https://petstore.swagger.io/v2/swagger.json",
"API_HEADERS":"Accept:application/json",
"API_WHITE_LIST":"addPet,updatePet,findPetsByStatus"
}
}
}
}
```
### Configuration
- List of available environment variables
- `DEBUG`: Enable debug logging (optional default is False)
- `OPENAPI_SPEC_PATH`: Path to the OpenAPI document. (required)
- `API_BASE_URL`: Base URL for the API requests. (required)
- `API_HEADERS`: Headers to include in the API requests (optional)
- `API_WHITE_LIST`: White Listed operationId in list format ["operationId1", "operationId2"] (optional)
- `API_BLACK_LIST`: Black Listed operationId in list format ["operationId3", "operationId4"] (optional)
- `HTTP_PROXY`: HTTP Proxy details (optional)
- `HTTPS_PROXY`: HTTPS Proxy details (optional)
- `NO_PROXY`: No Proxy details (optional)
## Contributing
Contributions are welcome.
Please feel free to submit a Pull Request.
## License
This project is licensed under the terms of the MIT license.
## Github Stars
[](https://star-history.com/#rahgadda/openapi_mcp_server&Date)
## Appendix
### UV
```bash
mkdir -m777 openapi_mcp_server
cd openapi_mcp_server
uv init
uv add mcp[cli] pydantic python-dotenv requests
uv add --dev twine setuptools
uv sync
uv run openapi_mcp_server
uv build
pip install --force-reinstall --no-deps .\dist\openapi_mcp_server-*fileversion*.whl
export TWINE_USERNAME="rahgadda"
export TWINE_USERNAME="<<API Key>>"
uv run twine upload --verbose dist/*
```
## Reference
- [UV Overview](https://www.youtube.com/watch?v=WKc2BdgmGZE)