Opik MCP

AIstdio
Repository

Comet's official MCP server for Opik, an open-source LLM evaluation and observability platform. Connects an AI host directly to an Opik workspace so agents can read traces and spans, log evaluation scores, manage prompt versions in the prompt library, and inspect project metrics without leaving the editor.

Connect

Add this configuration to .claude/mcp.json

{
  "mcpServers": {
    "opik": {
      "command": "npx",
      "args": [],
      "env": {
        "OPIK_API_KEY": "<YOUR_OPIK_API_KEY>",
        "OPIK_WORKSPACE": "<YOUR_OPIK_WORKSPACE>"
      }
    }
  }
}

Tools (3)

list_traces

List traces for an Opik project, with optional filters.

{
  "type": "object",
  "required": [
    "project_name"
  ],
  "properties": {
    "limit": {
      "type": "integer",
      "description": "Maximum number of traces to return"
    },
    "project_name": {
      "type": "string",
      "description": "Opik project name"
    }
  }
}

get_prompt

Retrieve a prompt and its versions from the Opik prompt library.

{
  "type": "object",
  "required": [
    "name"
  ],
  "properties": {
    "name": {
      "type": "string",
      "description": "Prompt name"
    }
  }
}

log_feedback_score

Attach a feedback or evaluation score to a trace.

{
  "type": "object",
  "required": [
    "trace_id",
    "name",
    "value"
  ],
  "properties": {
    "name": {
      "type": "string",
      "description": "Score name"
    },
    "value": {
      "type": "number",
      "description": "Score value"
    },
    "trace_id": {
      "type": "string",
      "description": "Target trace identifier"
    }
  }
}

Resources

This server does not expose any resources.

Prompts

This server does not expose any prompts.

Server information

Author
Comet ML (@comet-ml)
Repository
https://github.com/comet-ml/opik-mcp
License
Apache-2.0
Stars
0
Last updated
August 3, 2026