CodaMcpServer

The CodaMcpServer is an MCP server that provides a set of tools to interact with Coda documents, tables, and rows via the MCP protocol. It exposes endpoints to list documents, fetch tables, retrieve and update rows, and obtain metadata about columns. After starting the server, you can connect to it using any MCP-compliant client or the Cursor integration, enabling you to run tools like fetch-coda-docs, fetch-coda-tables, fetch-coda-table, fetch-coda-table-columns, fetch-coda-rows, upsert-coda-rows, and delete-coda-rows. The server authenticates requests using a Coda API key provided in the environment and exposes REST-like endpoints over SSE for real-time updates and a /messages endpoint for command handling.

Prerequisites:

• Node.js v20+ installed on your machine
• npm (comes with Node.js)
• A valid Coda API key

Installation steps:

1. Clone or download the repository to your local machine.
2. Navigate to the project directory.
3. Create a .env file in the project root with the required environment variables:

PORT=8000
CODA_API_KEY=your_coda_api_key_here

4. Install dependencies:

npm install

5. Build the TypeScript sources to JavaScript:

npx tsc

6. Run the server:

node dist/index.js

Optional combined startup (install, build, and run in one line):

npm install && npx tsc && node dist/index.js

The server will listen on port 8000 by default and expose endpoints such as GET /sse and POST /messages.

Notes and recommendations:

• Ensure your CODA_API_KEY is kept secret and available in the environment where the server runs.
• If you update TypeScript sources, re-run the build step (npx tsc) before starting the server.
• The tools support common parameters like docId, tableId, and row data structures; validate inputs using the server's runtime checks.
• Consider enabling pagination and caching for large datasets as mentioned in Known Limitations.
• If you encounter rate-limiting from the Coda API, implement retry logic or batching where supported by the client tooling.
• For deployment, you can adapt the mcp_config to a different runner (e.g., docker or uvx) if your environment requires it.