langfuse-mcp
MCP server for Langfuse that lets Claude Code create projects, manage API keys, and query traces without touching the web UI. Designed to work with a self-hosted Langfuse v3 stack running locally via Docker Compose.
README
langfuse-mcp
MCP server for Langfuse — lets Claude Code create projects, manage API keys, and query traces without touching the web UI.
Designed to work with langfuse-kickstart, a self-hosted Langfuse v3 stack running locally via Docker Compose.
Tools
| Tool | Description |
|---|---|
list_projects |
List all projects accessible with the current API key |
list_organizations |
List all organizations |
create_project |
Create a new project inside an organization |
create_api_key |
Create a new pk/sk pair for a project |
list_api_keys |
List API keys for a project |
delete_api_key |
Delete an API key by ID |
list_traces |
List traces with optional filters |
get_trace |
Get full trace details including observations |
list_sessions |
List sessions |
list_scores |
List scores with optional filters |
create_score |
Create a score on a trace or observation |
list_observations |
List spans, generations, and events |
list_datasets |
List datasets |
get_dataset |
Get a dataset and its items |
list_prompts |
List prompts in the prompt library |
get_prompt |
Get a prompt by name, version, or label |
Requirements
- Python 3.11+
- uv
- A running Langfuse instance (see langfuse-kickstart)
Setup
git clone https://github.com/dominic-righthere/langfuse-mcp
cd langfuse-mcp
cp .env.example .env
Edit .env:
LANGFUSE_HOST=http://langfuse.localhost
LANGFUSE_PUBLIC_KEY=pk-lf-kickstart
LANGFUSE_SECRET_KEY=sk-lf-kickstart
LANGFUSE_DATABASE_URL=postgresql://langfuse:YOUR_POSTGRES_PASSWORD@localhost:5432/langfuse
LANGFUSE_DATABASE_URL is needed for admin tools (create_project, create_api_key, etc.) that write directly to Postgres. With langfuse-kickstart, Postgres is exposed on 127.0.0.1:5432 — use the POSTGRES_PASSWORD from your kickstart .env.
Claude Code integration
Add to your .mcp.json (project-level) or claude_desktop_config.json:
{
"mcpServers": {
"langfuse": {
"command": "uv",
"args": ["run", "--project", "/path/to/langfuse-mcp", "langfuse-mcp"],
"env": {
"LANGFUSE_HOST": "http://langfuse.localhost",
"LANGFUSE_PUBLIC_KEY": "pk-lf-kickstart",
"LANGFUSE_SECRET_KEY": "sk-lf-kickstart",
"LANGFUSE_DATABASE_URL": "postgresql://langfuse:YOUR_POSTGRES_PASSWORD@localhost:5432/langfuse"
}
}
}
}
Replace /path/to/langfuse-mcp with the actual path to the cloned repo.
Alternatively, if you use a .env file in the repo root, you can omit the env block and let the server load it automatically.
Notes
create_api_keyreturns the secret key only at creation time — store it immediately.- Admin tools (
create_project,create_api_key,delete_api_key,list_organizations) connect directly to Postgres. If Langfuse changes its schema in a major version these tools may need updating. - The REST API tools (
list_traces,get_trace, etc.) use the public Langfuse API and are schema-stable.
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.