Skip to content
Langfuse logo

Langfuse MCP server

Which MCP servers are Langfuse's own, from the official registry and from asking its own domain, whether they answer, and what they offer. A program checked each of these, nobody copied them from a product page.

read 2026-10-08

← Back to Langfuse

Whose server it is
  • Addresshttps://langfuse.com/api/mcpnot in the official registry: answered when asked

No entry in the official MCP registry names this server. It answered on langfuse.com when the Atlas asked at the address servers are usually mounted on, so it is the vendor's own. Clients and catalogues that list servers from the registry do not show it.

When we asked itchecked 2026-10-26
  • Stateanswerschecked 2026-10-26
  • Tools4
  • Server calls itselfmcp-typescript server on vercel
  • Server version0.1.0
  • Protocol version2025-06-18

On 2026-10-26 the server answered initialize and tools/list without a sign-in. First answer in 95 ms.

Should an agent connect itSkip it
  • Its tools do none of the product's use cases we checked: trace LLM calls, tool steps and retrievals in agents; run evaluators and human review on production traces; version prompts and roll back changes without code deploys.
  • Checked on 6 of the 7 days it takes to judge reliability.

Once connected

  • Parameters are untyped or undescribed in 1 of 4 tools: check arguments before calling.

Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.

Measured dailydaily checks: 6, the first on 2026-10-27
  • INITE MCP scoreafter a weeka score needs 7 daily checks: 6 so far
  • What an agent can do with it0 of 3 use cases
  • Answeredmeasuring6 of the 7 daily checks it takes to print a share
  • Median first answer123 ms
  • Space in an agent's context≈ 934 tokensthe definitions of all its tools (4), which an agent carries on every turn
  • Tool definitionsA · 92 of 100
  • Typed results0 of 4 toolsdeclared outputSchema: results come typed, read once
  • Parameters without a description: in 1 of 4
  • Not saying whether it only reads or can destroy: 1 of 4

What an agent can do with it

  • cannot · trace LLM calls, tool steps and retrievals in agents
  • cannot · run evaluators and human review on production traces
  • cannot · version prompts and roll back changes without code deploys

Each of the product's use cases, put to Jev against this server's tool list: can an agent do it with these tools?

Worth a look

None of the five warning signs in these definitions.

A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (25%), doing the product's job (25%), definitions (20%), context (10%), use (10%) and how readily a client connects (10%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →

Tools it offers4 tools · read 2026-10-26
  • submitFeedback

    Submit explicit user-approved feedback to the Langfuse team. Before calling, ask permission and show the exact payload. If the user wants a reply, ask them to include their email address in the feedback text; only use an address they…

  • searchLangfuseDocs

    Semantic search (RAG) over the Langfuse documentation. Use this whenever the user asks a broader question that cannot be answered by a specific single page. Returns a concise answer synthesized from relevant docs. The raw provider…

  • getLangfuseDocsPage

    Fetch the raw Markdown for a single Langfuse docs page. Accepts a docs path (e.g., /docs/observability/overview) or a full https://langfuse.com URL. Returns the exact Markdown (may include front matter). Use when you need a specific page…

  • getLangfuseOverview

    Get a high-level, machine-readable index by downloading https://langfuse.com/llms.txt. Use this at the start of a session when needed to discover key docs endpoints or to seed follow-up calls to searchLangfuseDocs or getLangfuseDocsPage…

How to connect

Add the server to your MCP client config. Claude Desktop, Cursor and most other clients read this shape:

{
  "mcpServers": {
    "langfuse": {
      "type": "http",
      "url": "https://langfuse.com/api/mcp"
    }
  }
}

Add it

Claude Code, in a terminal

claude mcp add --transport http langfuse https://langfuse.com/api/mcp

Add to CursorAdd to VS Code

claude.ai, ChatGPT and other clients: add https://langfuse.com/api/mcp as a custom connector in the settings.

Each line is something a program found when it went looking: the registry entry, the server's own answer, a spec that parses. A missing line means we did not find it, not that it does not exist.