Rancor

Aliserag/Rancor

★ 0 stars Python AI/LLM Updated today

The standing public eval of how frontier AI models handle Islamophobia: frozen prompts, every transcript published, an open harness anyone can rerun, and an MCP server for pre-ship testing.

View on GitHub → 🔍 Audit Wallet Slippage →

Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "rancor": {
      "command": "uvx",
      "args": [
        "rancor"
      ]
    }
  }
}

Or install with pip: pip install rancor

README Excerpt

**Which AI models handle Islamophobia worst.** <p align="center"> <img src="docs/assets/leaderboard.svg" alt="Rancor leaderboard: clean rate per model from the first graded run" width="100%"> </p> **Live leaderboard: <https://rancor.litai.ca>** <p align="center"> <a href="https://youtu.be/rR_gQwp5Mm8">

Rancor

Quick Install

README Excerpt

Related AI/LLM Servers