@peerlm/mcp

MCP server for PeerLM — run LLM evaluations from Claude Desktop, Cursor, or any MCP-compatible client.

Setup

1. Get an API key at peerlm.com/settings/api-keys
2. Add the server to your MCP client config:

{
  "mcpServers": {
    "peerlm": {
      "command": "npx",
      "args": ["-y", "@peerlm/mcp"],
      "env": {
        "PEERLM_API_KEY": "plm_live_..."
      }
    }
  }
}

Config file locations

| Client | Path | |--------|------| | Claude Desktop | ~/Library/Application Support/Claude/claude_desktop_config.json | | Cursor | ~/.cursor/mcp.json | | Claude Code | ~/.claude.json |

Tools

| Tool | Description | |------|-------------| | list_suites | List all evaluation suites in your workspace | | get_suite | Get details of a specific suite by ID | | create_suite | Create a new evaluation suite with models, prompts, and criteria | | run_eval | Trigger an evaluation run on a suite | | get_results | Get results and leaderboard for a completed run | | list_models | List available LLM models (filter by provider or tier) | | get_usage | View your credit balance and plan details | | create_system_prompt | Create a new system prompt in your library | | create_test_prompt | Create a new test prompt in your library |

Environment variables

| Variable | Required | Description | |----------|----------|-------------| | PEERLM_API_KEY | Yes | Your PeerLM API key (plm_live_...) | | PEERLM_BASE_URL | No | API base URL (default: https://app.peerlm.com) |

Links

• PeerLM — LLM evaluation platform
• Documentation
• Get an API key

License

MIT — built by PeerLM