What is Langfuse Mcp

Langfuse MCP is a Model Context Protocol server designed for Langfuse, enabling AI agents to query trace data for enhanced debugging and observability.

Use cases

Use cases for langfuse-mcp include debugging AI models, tracking user sessions, monitoring application errors, and analyzing user activity for better insights.

How to use

To use langfuse-mcp, install the required dependencies, particularly ‘uv’, and utilize the provided tools to fetch traces, observations, and manage sessions and exceptions.

Key features

Key features include integration with Langfuse for trace data, a suite of tools for querying trace information, exception and error tracking, and session/user activity monitoring.

Where to use

Langfuse MCP can be utilized in software development, AI debugging, and any environment where monitoring and analyzing trace data is essential for improving application performance.

Langfuse MCP (Model Context Protocol)

This project provides a Model Context Protocol (MCP) server for Langfuse, allowing AI agents to query Langfuse trace data for better debugging and observability.

Quick Start with Cursor

Installation Options

🎯 From Cursor IDE: Click the button above (works seamlessly!)
🌐 From GitHub Web: Copy this deeplink and paste into your browser address bar:

cursor://anysphere.cursor-deeplink/mcp/install?name=langfuse-mcp&config=eyJjb21tYW5kIjoidXZ4IiwiYXJncyI6WyJsYW5nZnVzZS1tY3AiLCItLXB1YmxpYy1rZXkiLCJZT1VSX1BVQkxJQ19LRVkiLCItLXNlY3JldC1rZXkiLCJZT1VSX1NFQ1JFVF9LRVkiLCItLWhvc3QiLCJodHRwczovL2Nsb3VkLmxhbmdmdXNlLmNvbSJdfQ==

⚙️ Manual Setup: See Configuration section below

💡 Note: The “Add to Cursor” button only works from within Cursor IDE due to browser security restrictions on custom protocols (cursor://). This is normal and expected behavior per Cursor’s documentation.

After installation: Replace YOUR_PUBLIC_KEY and YOUR_SECRET_KEY with your actual Langfuse credentials in Cursor’s MCP settings.

Features

• Integration with Langfuse for trace and observation data
• Tool suite for AI agents to query trace data
• Exception and error tracking capabilities
• Session and user activity monitoring

Available Tools

The MCP server provides the following tools for AI agents:

• fetch_traces - Find traces based on criteria like user ID, session ID, etc.
• fetch_trace - Get a specific trace by ID
• fetch_observations - Get observations filtered by type
• fetch_observation - Get a specific observation by ID
• fetch_sessions - List sessions in the current project
• get_session_details - Get detailed information about a session
• get_user_sessions - Get all sessions for a user
• find_exceptions - Find exceptions and errors in traces
• find_exceptions_in_file - Find exceptions in a specific file
• get_exception_details - Get detailed information about an exception
• get_error_count - Get the count of errors
• get_data_schema - Get schema information for the data structures

Setup

Install uv

First, make sure uv is installed. For installation instructions, see the uv installation docs.

If you already have an older version of uv installed, you might need to update it with uv self update.

Installation

uv pip install langfuse-mcp

Obtain Langfuse credentials

You’ll need your Langfuse credentials:

• Public key
• Secret key
• Host URL (usually https://cloud.langfuse.com or your self-hosted URL)

Running the Server

Run the server using uvx:

uvx langfuse-mcp --public-key YOUR_KEY --secret-key YOUR_SECRET --host https://cloud.langfuse.com

Configuration with MCP clients

Configure for Cursor

Create a .cursor/mcp.json file in your project root:

{
  "mcpServers": {
    "langfuse": {
      "command": "uvx",
      "args": [
        "langfuse-mcp",
        "--public-key",
        "YOUR_KEY",
        "--secret-key",
        "YOUR_SECRET",
        "--host",
        "https://cloud.langfuse.com"
      ]
    }
  }
}

Configure for Claude Desktop

Add to your Claude settings:

{
  "command": [
    "uvx"
  ],
  "args": [
    "langfuse-mcp"
  ],
  "type": "stdio",
  "env": {
    "LANGFUSE_PUBLIC_KEY": "YOUR_KEY",
    "LANGFUSE_SECRET_KEY": "YOUR_SECRET",
    "LANGFUSE_HOST": "https://cloud.langfuse.com"
  }
}

Output Modes

Each tool supports different output modes to control the level of detail in responses:

• compact (default): Returns a summary with large values truncated
• full_json_string: Returns the complete data as a JSON string
• full_json_file: Saves the complete data to a file and returns a summary with file information

Development

Clone the repository

git clone https://github.com/yourusername/langfuse-mcp.git
cd langfuse-mcp

Create a virtual environment and install dependencies

uv venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate
uv pip install -e ".[dev]"

Set up environment variables

export LANGFUSE_SECRET_KEY="your-secret-key"
export LANGFUSE_PUBLIC_KEY="your-public-key"
export LANGFUSE_HOST="https://cloud.langfuse.com"  # Or your self-hosted URL

Testing

To run the demo client:

uv run examples/langfuse_client_demo.py --public-key YOUR_PUBLIC_KEY --secret-key YOUR_SECRET_KEY

Or use the convenience wrapper:

uv run run_mcp.py

Version Management

This project uses dynamic versioning based on Git tags:

1. The version is automatically determined from git tags using uv-dynamic-versioning
2. To create a new release:
   • Tag your commit with git tag v0.1.2 (following semantic versioning)
   • Push the tag with git push --tags
   • Create a GitHub release from the tag
3. The GitHub workflow will automatically build and publish the package with the correct version to PyPI

For a detailed history of changes, please see the CHANGELOG.md file.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Cache Management

We use the cachetools library to implement efficient caching with proper size limits:

• Uses cachetools.LRUCache for better reliability
• Configurable cache size via the CACHE_SIZE constant
• Automatically evicts the least recently used items when caches exceed their size limits