SIGN UPLOG IN
X (Twitter)LinkedInGitHubDiscordYouTube

Langfuse MCP server for AI agents and assistants

Securely connect your AI agents and chatbots (Claude, ChatGPT, Cursor, etc) with Langfuse MCP or direct API to inspect traces, review evals, manage prompts, and analyze metrics through natural language.

Get started for freeGet a demo
Langfuse logoLangfuse
Api Key

Langfuse is an open source LLM engineering platform for traces, evals, prompt management, and metrics. It helps teams debug, monitor, and improve LLM applications with clear observability data.

62 Tools

Try Langfuse now

Type what you want done — sign in and watch it run live in the Tool Router playground.

TOOL ROUTER PLAYGROUND
Langfuse
Try asking
TOOLS

Supported Tools

Every Langfuse action and event your agent gets out of the box.

Add Annotation Queue Item

Add a trace, observation, or session to a Langfuse annotation queue by queue ID and object ID.

Assign Annotation Queue User

Assign a user to a Langfuse annotation queue by queue ID and user ID.

Check Health

Check the health of the Langfuse API and database.

Create annotation queue

Create an annotation queue in the authenticated Langfuse project for human evaluation, using the supplied score configuration IDs.

Create comment

Create a comment on an existing Langfuse trace, observation, session, or prompt in the specified project.

Create Dataset

Create a dataset in the connected Langfuse project, optionally with a description, metadata, and JSON Schemas for item inputs and expected outputs.

Create Dataset Item

Create a Langfuse dataset item with input, expected output, and metadata.

Create Evaluation Rule

Create a Langfuse evaluation rule for incoming observations, using stable evaluator IDs and their latest versions.

Create evaluator

Create a version-1 evaluator in the authenticated Langfuse project.

Create media upload URL

Obtain a presigned upload URL and media ID for Langfuse media associated with either a trace (optionally an observation) or a dataset item.

Create model

Create a custom Langfuse model definition for usage-to-USD cost calculation and optional tokenization.

Create Prompt

Create a new version of a named Langfuse text or chat prompt.

Create Score

Attach an evaluation score to exactly one Langfuse trace, session, or dataset run, optionally narrowing a trace score to an observation.

Create score config

Create a score configuration in the authenticated Langfuse project to define numeric bounds, categorical label/value pairs, boolean scores, or text scores.

Delete Dataset Item

Delete a Langfuse dataset item by its ID and all its run items.

Delete Evaluation Rule

Delete a Langfuse evaluation rule by its stable ID, including legacy trace and dataset rules.

Delete Evaluator

Delete a Langfuse evaluator and all of its stored versions by its stable ID.

Delete LLM Connection

Delete a Langfuse LLM connection by its ID.

Delete Model

Delete a project-owned Langfuse model by its ID.

Delete Prompt

Delete Langfuse prompt versions by prompt name.

Delete Score

Delete a Langfuse score by its ID.

Delete Trace

Delete a specific Langfuse trace by its unique trace ID.

Delete Traces

Delete multiple Langfuse traces by their trace IDs.

Get Annotation Queue Item

Get a specific item from a Langfuse annotation queue by queue ID and queue item ID.

Get comment

Retrieve one comment by its unique Langfuse ID in the authenticated project.

Get Dataset

Retrieve a Langfuse dataset by name, including its metadata and JSON Schemas for dataset item inputs and expected outputs.

Get Dataset Item

Retrieve a single Langfuse dataset item by its ID, including its input, expected output, metadata, source IDs, and resolved media references.

Get Evaluation Rule

Get one Langfuse evaluation rule by its stable identifier, including legacy trace or dataset rules.

Get evaluator

Get one evaluator by its stable identifier in the authenticated Langfuse project.

Get media

Retrieve one media record by its unique Langfuse ID in the authenticated project.

Get model

Retrieve one Langfuse-managed or project-custom model definition by ID.

Get Project

Get the project associated with the connected project-scoped API key, returned in a data list with its organization, metadata, and retention configuration.

Get Prompt

Retrieve a Langfuse text or chat prompt by name, optionally selecting a deployment label or exact version and resolving dependencies.

Get score configuration

Retrieve one score configuration by ID in the authenticated Langfuse project.

Ingest Traces

Export OpenTelemetry traces to Langfuse using OTLP HTTP/JSON with real-time v4 ingestion.

List Annotation Queue Items

List one page of items in a Langfuse annotation queue, optionally filtered by PENDING or COMPLETED status.

List Annotation Queues

List annotation queues configured for human evaluation in the connected Langfuse project.

List comments

List one page of comments in the authenticated Langfuse project.

List Dataset Items

List dataset items, optionally scoped to a dataset, source trace or observation, or historical dataset version.

List Datasets

List datasets available to the connected Langfuse project.

List evaluation rules

List one page of evaluation rules in the authenticated Langfuse project, newest creation first.

List evaluators

List one page of evaluators in the authenticated Langfuse project, newest creation first.

List evaluator versions

List one page of an evaluator's version history, newest first, for history and audit use cases.

List Experiment Items

List Langfuse experiment items started on or after a required datetime, ordered by time descending.

List Experiments

List Langfuse experiments active from a required start time, optionally specifying an end time and including metadata and directly attached scores.

List LLM connections

List one page of LLM provider connections in the authenticated Langfuse project, including model configuration and masked secrets.

List Models

List Langfuse-managed and project-custom model pricing definitions used for usage and cost calculation.

List Observations

Search Langfuse observations such as generations, spans, events, agents, and tool calls, selecting only the field groups needed.

List prompts

List one page of prompt names, types, versions, labels, tags and latest matching configuration in the authenticated Langfuse project.

List score configurations

List one page of score configurations in the authenticated Langfuse project, including score types, numeric bounds, categorical labels, and archive status.

List Scores

Search numeric, boolean, categorical, text, and correction scores with subject and annotation context.

Query Metrics

Run an aggregate Langfuse metrics query over observations or numeric, boolean, or categorical scores.

Remove Annotation Queue Assignment

Remove a user's assignment to a Langfuse annotation queue by queue ID and user ID.

Remove Annotation Queue Item

Remove an item from a Langfuse annotation queue using the queue ID and annotation queue item ID.

Submit feedback

Submit feedback to Langfuse about its skills, MCP tools, CLI, docs, or public API.

Update Annotation Queue Item

Update a Langfuse annotation queue item's status to PENDING or COMPLETED using its queue ID and item ID.

Update Evaluation Rule

Update a Langfuse evaluation rule by its stable ID.

Update evaluator

Update a Langfuse evaluator by stable identifier.

Update media

Update an existing Langfuse media record with its upload timestamp and HTTP status, and optionally the upload error message and elapsed time.

Update Prompt Labels

Update deployment labels for a specific Langfuse prompt version by name and version.

Update score config

Update an existing Langfuse score configuration by config ID.

Upsert Model

Create or replace a project-owned Langfuse model definition by ID for usage cost calculation and optional tokenization.

SETUP GUIDE

Connect Langfuse MCP Tool with your Agent

1

Install Composio

typescript
npm install @composio/core ai @ai-sdk/mcp @ai-sdk/openai
Install the Composio SDK and your agent framework
2

Create a session with MCP enabled

typescript
import { Composio } from "@composio/core";

const composio = new Composio();
const { mcp } = await composio.create("your-user-id", {
  toolkits: ["langfuse"],
  mcp: true,
});
Create a session scoped to Langfuse and read its MCP URL and headers
3

Connect your agent to the MCP server

typescript
import { createMCPClient } from "@ai-sdk/mcp";
import { openai } from "@ai-sdk/openai";
import { generateText, stepCountIs } from "ai";

const client = await createMCPClient({
  transport: { type: "http", url: mcp.url, headers: mcp.headers },
});

const { text } = await generateText({
  model: openai("gpt-5.6-sol"),
  tools: await client.tools(),
  prompt: "List recent Langfuse traces with errors and summarize common failure patterns",
  stopWhen: stepCountIs(10),
});

console.log(text);
await client.close();
Pass the session's MCP URL and headers to your agent and run a Langfuse request
SETUP GUIDE

Connect Langfuse API Tool with your Agent

1

Install Composio

typescript
npm install @composio/core @composio/openai openai
Install the Composio SDK, the OpenAI provider, and the OpenAI SDK
2

Create a Composio session

typescript
import OpenAI from "openai";
import { Composio } from "@composio/core";
import { OpenAIResponsesProvider } from "@composio/openai";

const composio = new Composio({ provider: new OpenAIResponsesProvider() });
const client = new OpenAI();

const session = await composio.create("your-user-id", { toolkits: ["langfuse"] });
const tools = await session.tools();
Initialize Composio with the OpenAI Responses provider and create a session scoped to Langfuse
3

Run Langfuse tools with your agent

typescript
let response = await client.responses.create({
  model: "gpt-5.6-sol",
  tools,
  input: [{ role: "user", content: "List recent Langfuse traces with errors and summarize common failure patterns" }],
});

while (response.output.some((o) => o.type === "function_call")) {
  const outputs = await composio.provider.handleToolCalls(session, response.output);
  response = await client.responses.create({
    model: "gpt-5.6-sol",
    tools,
    previous_response_id: response.id,
    input: outputs,
  });
}

console.log(response.output_text);
Send a request, execute the Langfuse tool calls through the session, and print the final answer

Why Use Composio?

AI Native Langfuse Integration

  • Supports both Langfuse MCP and direct API based integrations
  • Structured, LLM-friendly schemas for reliable trace, eval, prompt, and metric workflows
  • Rich coverage for reading, writing, and querying your Langfuse observability data

Managed Auth

  • Secure API key handling so you don't have to hard-code Langfuse credentials
  • Central place to manage, scope, and revoke Langfuse access
  • Per user and per environment credentials for safer agent deployments

Agent Optimized Design

  • Tools are tuned for AI agents, so they can work with Langfuse traces and metrics more reliably
  • Clear execution logs show what your agent ran, when it ran, and on whose behalf
  • Less glue code, fewer brittle integrations, and faster debugging loops

Enterprise Grade Security

  • Fine-grained RBAC so you control which agents and users can access Langfuse
  • Scoped, least privilege access to Langfuse resources
  • Full audit trail of agent actions to support review, security checks, and compliance
FAQ

Frequently asked questions

Yes, Langfuse requires you to configure your own API key. Once set up, Composio handles secure credential storage and API request handling for you.

Yes! Composio's Tool Router enables agents to use multiple toolkits. Learn more.

Yes. Composio is SOC 2 Type II compliant and is built to keep your Langfuse connection and credentials secure. OAuth tokens and API keys are encrypted, and sensitive customer data is protected at rest and in transit.

Composio also undergoes independent security testing and continuously monitors its systems for security threats. You can review the latest reports and policies in the Composio Trust Center.

Composio maintains and updates all toolkit integrations automatically, so your agents always work with the latest API versions.

Create the key in your Langfuse account settings, then paste it once on Composio's connection page. Composio stores it encrypted and uses it only for the Langfuse actions your agent runs. Nobody else in your workspace can read it, and you can revoke it inside Langfuse at any time.

When you connect a Langfuse account through Composio, every action runs under that account, so anything the agent creates, sends, or changes shows up in Langfuse as done by you. Keep an approval step in your prompt for actions with side effects, such as sending or deleting, and have the agent draft first.

Whatever the credentials you connect allow inside Langfuse. If Langfuse lets you scope a key to specific permissions, create a scoped one so the agent can only do what you intend. You can revoke the key inside Langfuse at any time.

Yes. Composio supports multiple connected accounts for the same app, and that works in Claude, ChatGPT, or any other assistant you connect through Composio. Give each Langfuse connection a name such as work or personal, and the assistant uses the one you mention in the request. Each account keeps its own credentials and nothing is merged.

The connection stops working the moment Langfuse rejects the old key. Create a new key in Langfuse and reconnect the account from the Composio dashboard or by asking your agent to reconnect Langfuse. Nothing else changes.

Composio's free Hobby plan includes 100,000 tool calls per month with no credit card, which covers most personal Langfuse use. Paid plans add higher limits and team features. Your Langfuse plan and its API limits still apply as usual.

Built-in connectors usually give one AI access to a limited set of apps. Many people use Composio because it lets their AI connect to more apps than it normally supports, or connect to multiple accounts for the same app (e.g. connect Claude to multiple Langfuse accounts).

With Composio, you connect Langfuse once and then use it across different AI assistants without setting it up separately in each one. Connect your apps to Composio once, then connect Composio to whichever AI you use, whether that's Claude, ChatGPT, Hermes, or your own custom assistant.

Start with Langfuse.It takes 30 seconds.

Managed auth, hosted MCP servers, and every Langfuse tool your agent needs.Free to start.

Start building