home / mcp / backlog mcp server

Backlog MCP Server

Provides a Backlog data bridge with controlled MCP endpoints for reading backlog items and related actions.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "kazuhideoki-mcp-server-backlog": {
      "command": "node",
      "args": [
        "/path/to/mcp-server-backlog/dist/index.js",
        "--permission",
        "READ"
      ]
    }
  }
}

This MCP server provides a Backlog integration you can run locally or in your environment. It lets you access Backlog data and perform actions through a controlled MCP interface, with a simple setup and clear permission levels to protect your data.

How to use

Prepare your Backlog connection by placing your API key in a file named apikey at the project root. Then install dependencies, build the server, and integrate it into your MCP client workflow. You can run the server in development or production mode, and you control access with a permission flag to limit what endpoints you can call.

Next, add the MCP server to your Claude-based setup so your client can talk to it. You will reference the local built file that serves the MCP endpoint.

claude mcp add backlog node /path/to/mcp-server-backlog/dist/index.js -- --permission READ

How to install

Prerequisites: ensure you have Node.js and npm installed on your machine.

Install dependencies and build the server:

npm install
npm run build

Run the MCP server

You can run the server in development or production mode. Use a permission level to control access to endpoints.

# Development mode with permission flag
npm run dev -- --permission READ

# Production mode with permission flag (after build)
node /path/to/mcp-server-backlog/dist/index.js --permission READ

Permission Levels

- READ: Read-only access (can only use GET endpoints) - MUTATE: Full access (can use all endpoints including POST, PUT, PATCH, DELETE)

Notes

- The Backlog MCP server expects the API key to be available in a file named apikey at the project root. - The primary run path for a built server is the node command targeting the built index in dist/index.js with a --permission flag.

Security and configuration tips

- Use a dedicated API key with scoped permissions for MCP usage to minimize risk. - Keep the apikey file out of version control by adding it to your ignore list or environment-specific vaults. - When deploying, prefer production mode with an explicit permission level to avoid accidental data exposure.

Available tools

devServer

Launches the local MCP server for development with a permission flag to restrict access during testing.

buildProcess

Builds the MCP server into a distributable dist folder for production usage.