Forgejudge

Overview

Forgejudge is an MCP server that provides an open evaluation leaderboard and CI gate for autonomous coding agents. It lets you solve problems, score results, and trace execution paths for agent-based coding workflows. The server exposes these capabilities as tools that Claude, Cursor, and other AI assistants can invoke directly.

It runs locally on your machine via the Python package forgejudge, installed with uvx forgejudge. The server communicates over stdio and requires no authentication or credentials to get started. The latest release is v0.1.1, licensed under MIT, with source code available at ahmedEid1/forgejudge on GitHub.

Forgejudge sits in the AI & Agents category and works with MCP clients including Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, Zed, and others. If you maintain this project, claim your listing on mcp.site to verify ownership and keep details up to date.

Installation

// claude_desktop_config.json
{
  "mcpServers": {
    "forgejudge": {
      "command": "uvx",
      "args": [
        "forgejudge"
      ]
    }
  }
}