mcp-devbrain-stdio

DevBrain MCP Server provides a newsletter-oriented knowledge access toolset. It searches for relevant code snippets, indie developer articles, and blog posts, returning concise results with descriptions and URLs that you can feed into your preferred LLM agent (e.g., Claude, Goose) to fetch full content or build features. You can deploy it in stdio mode via uvx or run a Docker container, and you can integrate it with Claude’s configuration to expose an mcpServers entry that points to the running instance. The README demonstrates both a local stdio workflow and a Docker-based workflow, as well as integration steps for Claude Code and Claude’s UI configuration. Tools available include a programmable MCP endpoint that returns compact article summaries and links, and the ability to direct an agent to fetch full article content from the provided URLs for downstream feature implementation.

Prerequisites:

• Python 3.10+ installed on the host
• uv/uvx tooling (for stdio mode)
• Docker installed (for docker mode, optional)

Install uv/uvx (stdio workflow):

curl -LsSf https://astral.sh/uv/install.sh | sh

Run the DevBrain MCP in stdio mode (stdio target):

uvx --python ">=3.10" --from devbrain devbrain-stdio-server

Alternatively, run via Docker (docker target):

docker run -i --rm svenai/mcp-devbrain-stdio:latest

If using Claude integration, configure Claude’s mcpServers with the stdio command shown above, for example:

{
  "mcpServers": {
    "devbrain": {
      "command": "uvx",
      "args": [
        "--python", ">=3.10",
        "--force-reinstall",
        "--from",
        "devbrain",
        "devbrain-stdio-server"
      ]
    }
  }
}

Tips and caveats:

• DevBrain requires Python 3.10+; ensure your environment uses a compatible Python interpreter. Some environments (e.g., VS Code shipped Python) may default to older versions; specify --python ">=3.10" explicitly.
• If Claude (or similar agents) reports issues with uv/uvx, follow the post-install linking steps provided in the README to ensure the binaries are accessible from /usr/local/bin.
• The Docker approach uses the image svenai/mcp-devbrain-stdio:latest; for persistent config across restarts, consider mounting a config volume or adapting your agent’s config to point to the running container.
• When using Claude or other agents, ensure you add the correct mcpServers entry to the agent’s configuration JSON and use the provided command/args exactly as shown to avoid version mismatches.
• The README also provides a prebuilt remote HTTP MCP example (devbrain type http) if you prefer a remote endpoint instead of local stdio or Docker.
• If you encounter version conflicts with Python within host tooling, verify PATH and PYTHONHOME, and consider using a dedicated virtual environment for DevBrain.