MCP Catalogs
Home

opik-mcp

by comet-ml·203·Score 50

Comet-ML's MCP server for Opik provides unified access to prompts, projects, traces, and metrics across various IDEs.

ai-llmdeveloper-toolsmonitoring
30
Forks
15
Open issues
this month
Last commit
2d ago
Indexed

Overview

This is a production-ready MCP server implementation that bridges Opik's ML experiment tracking platform with IDE environments via the Model Context Protocol. It supports both local stdio and remote streamable-http transports, making it versatile for different deployment scenarios. The server exposes comprehensive toolsets for managing prompts, projects, traces, metrics, and datasets, with fine-grained configuration options to enable specific capabilities based on user needs.

Try asking AI

After installing, here are 3 things you can ask your AI assistant:

you:ML engineers tracking experiment metrics and accessing prompts directly within their IDE
you:Data scientists exploring project traces and datasets without leaving their development environment
you:Teams managing prompt lifecycle and version control across different workspace projects

When to choose this

Choose this MCP server if you're already using or planning to use Opik for MLOps and want seamless IDE integration with prompts, projects, traces, and metrics.

When NOT to choose this

Avoid if you need vendor-agnostic solutions, as this locks you into Comet's Opik platform, or if you require a hosted remote MCP service (currently not provided).

Tools this server exposes

5 tools extracted from the README (low confidence)
  • resources/list

    Exposes static URIs for workspace information

  • resources/templates/list

    Exposes dynamic URI templates for projects

  • resources/read

    Reads data from static and templated URIs

  • prompts/list

    Lists available workflow prompts

  • prompts/get

    Retrieves specific prompts by ID

Note: Tool names were inferred from the documentation section about MCP resources and prompts, as no explicit tool list was provided in the README

Comparable tools

langsmith-mcpembedchain-mcpllamaindex-mcpsemantic-kernel-mcp

Installation

Installation

  1. Run with npx:
npx -y opik-mcp --apiKey YOUR_API_KEY
  1. Add to Claude Desktop (claude_desktop_config.json):
{
  "mcpServers": {
    "opik": {
      "command": "npx",
      "args": ["-y", "opik-mcp", "--apiKey", "YOUR_API_KEY"]
    }
  }
}
  1. Add to VS Code / GitHub Copilot (.vscode/mcp.json):
{
  "inputs": [
    {
      "type": "promptString",
      "id": "opik-api-key",
      "description": "Opik API Key",
      "password": true
    }
  ],
  "servers": {
    "opik-mcp": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "opik-mcp", "--apiKey", "${input:opik-api-key}"]
    }
  }
}

Compare opik-mcp with

GitHub →

Last updated · Auto-generated from public README + GitHub signals.