mcp-coroot

This MCP server integrates with the Coroot observability platform, exposing a suite of tools that let MCP clients monitor applications, analyze performance, search logs, and view traces within Coroot. The server is configured to run via uvx and exposes the mcp-coroot package, enabling real-time metrics collection, log analysis, distributed tracing, and infrastructure insights from your Coroot instance. You can authenticate with a username/password or a session cookie (SSO/MFA), and you can supply an API key if you only need data ingestion access.

To use, start the server with uvx mcp-coroot, pointing at your Coroot instance via the provided environment variables. Once running, your MCP client can request the 61 available tools organized across authentication, project management, monitoring, dashboards, integrations, configuration, and advanced features to interact with Coroot data. Typical workflows include monitoring application health, querying logs and traces, managing projects and dashboards, configuring integrations, and performing RCA and profiling tasks through the MCP interface.

Prerequisites:

• Python 3.8+ and uv (or uvx) installed, or use the uvx installation flow as described below.
• Access to a Coroot instance for testing.

Using uvx (recommended):

1. Install and run directly uvx mcp-coroot

Using pip (for local development):

1. Install the package pip install mcp-coroot
2. Run the server via uvx integration when appropriate (if packaged as a plugin)

Using Docker (alternative):

1. Run with environment variables configured for your Coroot instance docker run --rm -i
   -e COROOT_BASE_URL="http://localhost:8080"
   -e COROOT_USERNAME="admin"
   -e COROOT_PASSWORD="your-password"
   jamesbrink/mcp-coroot:latest

From Source:

1. Clone the repository git clone https://github.com/jamesbrink/mcp-coroot.git
2. Navigate to the project and install dependencies cd mcp-coroot uv sync --all-groups
3. Run the server uv run mcp-coroot

Tips and common considerations:

• Ensure COROOT_BASE_URL points to your Coroot instance and that the credentials provided have the necessary permissions.
• For SSO/MFA setups, prefer COROOT_SESSION_COOKIE instead of username/password.
• If using API key only for data ingestion, set COROOT_API_KEY and limit access accordingly; avoid exposing management endpoints with API keys.
• When running in Docker, pass environment variables for base URL, credentials, and any required cookies or API keys as shown in the README examples.
• The server exposes a broad set of tools; consult Coroot documentation for tool-specific usage and expected input/output formats.