Warpmetrics MCP Server

Connect AI assistants to your Warpmetrics AI agent telemetry via the Model Context Protocol.

Installation

npm install -g @warpmetrics/mcp

Related MCP server: Datadog MCP Server

Setup

1. Get your API key

Create an API key at warpmetrics.com/app/api-keys

2. Configure Claude Desktop

Add to your ~/.claude/claude_desktop_config.json:

{
  "mcpServers": {
    "warpmetrics": {
      "command": "warpmetrics-mcp",
      "env": {
        "WARPMETRICS_API_KEY": "wm_live_your_api_key_here"
      }
    }
  }
}

3. Restart Claude Desktop

The Warpmetrics tools will now be available.

Available Tools

Tools are loaded dynamically from the Warpmetrics API and stay automatically in sync.

View all available tools:

• Documentation: warpmetrics.com/docs/mcp

• CLI: npx @warpmetrics/mcp --list-tools

Example Prompts

Runs & Calls

• "How many runs did I have today?"

• "Show me the most expensive calls"

• "List recent failed runs"

• "Show details for run wm_run_01abc123"

Costs & Performance

• "What's my total LLM spend this week?"

• "What's the average latency for my code-review agent?"

• "Show me cost trends for the last 7 days"

Outcomes & Success Rates

• "What's the success rate for my code-review agent?"

• "Show me outcome statistics"

• "List all outcome classifications"

Environment Variables

Development

git clone https://github.com/warpmetrics/mcp.git
cd mcp
npm install

# Run locally
WARPMETRICS_API_KEY=wm_live_... node src/index.js

License

MIT