Coda

Provides access to Coda documents, tables, and rows via MCP with read and write capabilities.

rust

2

GitHub Stars

rust

Language

4 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "nkpar-coda-mcp": {
      "command": "./target/release/coda-mcp",
      "args": [],
      "env": {
        "RUST_LOG": "info",
        "CODA_BASE_URL": "https://coda.io/apis/v1",
        "CODA_API_TOKEN": "your_token_here"
      }
    }
  }
}

You run a Model Context Protocol (MCP) server to let AI assistants access and modify Coda documents, tables, and rows. This server bridges your Coda.io data with your preferred AI tooling, enabling create, read, update, and delete operations as part of automated workflows while handling authentication and rate limits.

How to use

You can use this MCP server with a client such as Claude Desktop or Claude Code by configuring an MCP connection and providing your Coda API token. The server exposes a local or containerized process that your client talks to, allowing operations like listing documents, reading page content, manipulating rows, and more. Ensure your API token has write permissions if you intend to perform write actions.

How to install

Prerequisites you need before installing: a Rust toolchain (for building from source) and, optionally, Docker for containerized usage.

Install the MCP server from crates.io using Cargo.

cargo install coda-mcp

If you prefer to build from source, clone the repository, build in release mode, and run the produced binary.

git clone https://github.com/nkpar/coda-mcp.git
cd coda-mcp
cargo build --release

Binary will be located at `./target/release/coda-mcp


You can also run the server using Docker without installing Rust tooling. Pull the pre-built image and run it with your token.

docker pull ghcr.io/nkpar/coda-mcp:latest


To start the container and expose the MCP endpoint to your client, set the CODA_API_TOKEN environment variable inside the container and pass it to the process.

Claude users can connect by configuring Claude to run the MCP command within Docker as shown in the provided snippets.

Additional setup tips and examples

Set up Claude Desktop with the MCP configuration by pointing to the local binary when running directly, or by using Docker as the command with the appropriate environment variable.

If you build from source, you can run the server directly from the release binary.

Example capabilities include: listing documents, getting documents and pages, listing tables and rows, and adding or updating rows, all while respecting Coda API rate limits.

Available tools

list_docs

List available documents in your Coda workspace.

get_doc

Retrieve details about a specific document.

search_docs

Search documents by name.

create_doc

Create a new Coda document (optionally assign folder, template, timezone).

delete_doc

Permanently delete a document.

list_pages

List all pages within a document.

get_page

Get the HTML content of a page.

list_tables

List tables in a document.

get_table

Get details about a specific table.

list_columns

List columns in a table.

get_rows

Retrieve rows with optional filtering.

get_row

Get a specific row by identifier.

add_row

Add a new row to a table.

update_row

Update an existing row.

delete_row

Delete a row.

list_formulas

List named formulas.

get_formula

Get the value of a named formula.

list_controls

List controls like buttons or sliders in a document.