Skip to main content
MCP server CX Codex

Openlayer — Claude Code MCP server: what it does, how to install

Connects coding assistants to the Openlayer evaluation platform. Useful for inspecting runtime traces and debugging failing checks.

Install into Claude Code · Platform CLI: remote server All methods and platforms ↓
claude mcp add --transport http mcp https://mcp.openlayer.com/mcp
Open on GitHub

What Openlayer does

This MCP server hooks coding agents into the hosted endpoint at https://mcp.openlayer.com/mcp. It allows the assistant to inspect execution traces and review active governance rules. Debugging model issues becomes structured.

The tool assists with generating openlayer.json and tests.json config files. You can use these tests to gate CI/CD pipelines. All write or delete actions pause for user confirmation. Workspaces stay protected against unintended edits.

Running the server requires an OPENLAYER_API_KEY. On the initial tool call, the agent prompts for OAuth authentication to pick a workspace. Self-hosted instances require setting OPENLAYER_BASE_URL.

What it helps with

  • Investigate failing model checks with the MCP fix-loop
  • Generate openlayer.json and tests.json config files for offline evals
  • Inspect execution traces and governance rules inside a workspace
  • Gate CI/CD deployment pipelines on test outcomes

Example prompts for the agent

  • Inspect the latest traces in the workspace and explain why the test failed
  • Draft an openlayer.json file to evaluate model outputs
  • Review active governance rules and current evaluation status

How to install Openlayer

Platform CLI: remote server

claude mcp add --transport http mcp https://mcp.openlayer.com/mcp

Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more

Questions about Openlayer

How do I install Openlayer in Claude Code?

Add the server by running claude mcp add --transport http openlayer https://mcp.openlayer.com/mcp. On the first run, authenticate via OAuth and choose your workspace.

What does the Openlayer MCP server do?

It connects the coding agent to Openlayer resources, allowing it to inspect traces, check evaluation status, and investigate failures. It also helps generate eval config files for deployment pipelines.

What prerequisites are needed to use this server?

You need an Openlayer account and an OPENLAYER_API_KEY environment variable. If you run a self-hosted deployment, set OPENLAYER_BASE_URL to your custom endpoint.

Similar MCP servers

All MCP servers for AI and ML →

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor