Taiga MCP Bridge

What is Taiga MCP Bridge?

Taiga MCP Bridge is an integration layer that connects the Taiga project management platform with the Model Context Protocol (MCP), enabling AI tools and workflows to interact with Taiga’s resources. It is a community fork of talhaorak/pytaiga-mcp with additional features and ongoing maintenance.

How to use Taiga MCP Bridge?

Install Python 3.12+ and uv, then clone the repository and run ./install.sh. Configure environment variables (TAIGA_API_URL, TAIGA_USERNAME, TAIGA_PASSWORD) either in a .env file or by passing them to Docker. Start the server with ./run.sh (stdio) or ./run.sh --sse (SSE). Use a client that supports MCP (e.g., Claude App with .mcp.json configuration) to invoke the provided tools.

Key features of Taiga MCP Bridge

• Full CRUD for projects, epics, user stories, tasks, issues, and sprints.
• Secure credential handling – passwords never appear in logs.
• Auto-authentication on startup via TAIGA_USERNAME and TAIGA_PASSWORD.
• Allowlist-based input validation prevents unexpected data.
• Response verbosity control (minimal, standard, full) reduces AI context usage.
• Supports both stdio and SSE transport modes.

Use cases of Taiga MCP Bridge

• Create and manage projects, epics, user stories, tasks, and issues through AI assistants.
• Track sprint progress and milestones via Milestones (sprints) tools.
• Assign and update work items programmatically with AI.
• Query detailed information about project artifacts for reporting.
• Manage project members and permissions without manual UI interaction.

FAQ from Taiga MCP Bridge

How does authentication work?

You can either set TAIGA_USERNAME and TAIGA_PASSWORD environment variables for automatic startup authentication (recommended), or manually call the login tool to obtain a session ID and include it in subsequent requests. Logout terminates the session.

What transport modes are supported?

Two modes are available: stdio (standard input/output, default for terminal clients) and SSE (Server-Sent Events, web-based with server push). Set via the --sse flag, the TAIGA_TRANSPORT environment variable, or in the .env file.

What dependencies are required?

Python 3.12 or higher and the uv package manager. The server can also be run via Docker using the pre-built image at ghcr.io/tetra-2023/pytaiga-mcp:latest.

Where is data stored?

The server does not store data locally; all CRUD operations are performed directly on the Taiga API instance configured via TAIGA_API_URL.

Can I control how much data is returned?

Yes, every tool accepts a verbosity parameter: minimal (core fields only), standard (common fields including version), or full (complete API response). This helps reduce AI context usage when listing many items.