概要
What is Backlog MCP サーバー?
Backlog MCP サーバー is an MCP server that provides access to the Backlog API, enabling project management, issue tracking, file operations, and other features. It is designed for users who want to interact with Backlog through an MCP-compatible agent such as Claude Desktop.
How to use Backlog MCP サーバー？
Obtain an API key from Backlog (Settings > API > Issue API key). Set environment variables BACKLOG_DOMAIN and BACKLOG_API_KEY in a .env file, or configure them in claude_desktop_config.json when using with Claude Desktop via NPX. Then run the server with npx -y @tmhr1850/backlog-mcp-server.
Key features of Backlog MCP サーバー
- Retrieve space and project information.
- Create, update, and search issues.
- Add comments to issues.
- Manage attachments and user data.
- Search issues by keyword and filters.
Use cases of Backlog MCP サーバー
- Automating issue creation from an AI assistant.
- Querying project status and issue details without leaving a chat interface.
- Updating existing issues with comments or status changes.
- Searching for issues across projects using keywords and filters.
FAQ from Backlog MCP サーバー
What are the prerequisites to use this server?
A Backlog API key and your Backlog domain (e.g., your-domain.backlog.com) are required. Optionally, Node.js and npm are needed for local execution.
How is data accessed and where does it live?
All data is accessed via the Backlog API. The server does not store data locally; it makes HTTP requests to your Backlog instance.
What authentication or transport does it use?
The server uses API key authentication, passed via the BACKLOG_API_KEY environment variable. Transport is HTTP/HTTPS through the Backlog API.
Are there any limits on the number of results?
The getIssues and searchIssues tools accept a count parameter, with a maximum of 100 items per request.
How can I test the server locally?
Run npm test for all tests, npm run test:watch for watch mode, or npm run test:coverage for coverage reports.
