Opik MCP Server is a Model Context Protocol (MCP) server that provides an interface for prompt lifecycle management, workspace exploration, metrics operations, and resource management for MCP-compatible clients.
From the registry: Interact with Opik prompts, traces, datasets and metrics through the Model Context Protocol.
Please install the `opik-mcp` MCP server into my current AI client (that's you).
Required prerequisites (do these first if not already done):
- **Opik API Key** — Get an API key from Opik Cloud (comet.com). Not required for self-hosted Opik. (https://www.comet.com/site/products/opik/)
Canonical MCP server config (stdio transport):
- command: `npx`
- args: ["-y","opik-mcp","--apiKey","<your-api-key>"]
Note: API key is passed as CLI argument, not env var. For self-hosted Opik, add '--apiUrl http://localhost:5173/api'. Supports stdio and streamable-http transports. Requires Node.js >= 20.11.0.
Add this MCP server to my current client's config in the correct format for you. If you need secrets or credentials I haven't provided, ASK me — do not invent values or leave raw placeholders. After adding it, tell me how to verify the server is connected.OPIK_API_KEYrequiredAPI key for authenticating with the Opik service.OPIK_TOOLSETSComma-separated list of enabled toolsets.STREAMABLE_HTTP_REQUIRE_AUTHFlag to require authentication for streamable HTTP transport.STREAMABLE_HTTP_VALIDATE_REMOTE_AUTHFlag to validate remote authentication for streamable HTTP transport.REMOTE_TOKEN_WORKSPACE_MAPJSON token-to-workspace map for remote authentication.STREAMABLE_HTTP_TRUST_WORKSPACE_HEADERSFlag to trust workspace headers in remote mode.AI Agents Framework with Self Reflection and MCP support