chemdraw

ChemDraw Server provides a FastAPI-based MCP API to convert between chemical names and SMILES, as well as process molecules (e.g., RDKit objects) and compare molecules. The service is exposed under the MCP path and offers an API key protected interface. Once running, you can send HTTP requests to the provided endpoints to perform name_to_smiles, smiles_to_name, smiles_to_rdkit, and mol_compare operations. The server also exposes a health check endpoint to monitor availability. Authentication is via a Bearer token you supply in the Authorization header, configured through an API_KEY in your environment or a .env file.

Prerequisites:

• Python 3.10
• Git
• uv (the CLI tool used to run the server; installed as described in the project docs)

Installation steps:

1. Clone the repository: git clone https://github.com/tom832/chemdraw-server.git
2. Change into the project directory: cd chemdraw-server
3. Initialize the environment and install dependencies using uv (as per the project setup): uv sync
4. Create a .env file (optional but recommended) and add your API key: API_KEY=your_api_key_here DOCS_ACCESS_TOKEN=
5. Start the server: uv run main_server.py

The API will be available at:

• API: http://localhost:1145/chemdraw/api/
• MCP: http://localhost:1145/chemdraw/mcp/

Notes and tips:

• Ensure Python 3.10 is used to satisfy dependencies.
• The MCP endpoints require an Authorization header with a Bearer token. Set API_KEY in your environment or in a .env file.
• If you rely on Prometheus metrics, ensure the prometheus-fastapi-instrumentator is available in your environment.
• The server paths for API and MCP are /chemdraw/api/ and /chemdraw/mcp/ respectively; you can test health with GET /chemdraw/api/health.
• If you modify config.py, review configurable options as described in the project configuration to adjust timeouts, logging, or authentication behavior.