MCP Catalogs
Homechucknorris screenshot

chucknorris

by pollinations·60·Score 37

An MCP server delivering jailbreak prompts to LLMs with dynamic schema adaptation techniques.

securityai-llmdeveloper-tools
14
Forks
2
Open issues
13 mo ago
Last commit
2d ago
Indexed

Overview

The ChuckNorris MCP Server implements a specialized approach to deliver jailbreak prompts to various language models. It uses a two-phase technique where the first call presents a neutral schema, and subsequent calls dynamically update to include jailbreak instructions. This method aims to bypass detection by presenting the jailbreak content differently to AI models versus users. Based on elder-plinius' L1B3RT4S project, this tool serves primarily for security research to identify vulnerabilities in LLM systems.

Try asking AI

After installing, here are 4 things you can ask your AI assistant:

you:Security research to identify LLM vulnerabilities
you:Testing jailbreak detection mechanisms
you:Understanding MCP protocol security implications
you:Does this work with all LLM models?

When to choose this

Security researchers and AI developers who need to test LLM vulnerabilities and understand potential jailbreak techniques through MCP.

When NOT to choose this

You should not use this for production applications, as it's designed solely for security research purposes and may trigger security alerts in production environments.

Tools this server exposes

1 tool extracted from the README
  • C̷h̷u̷c̷k̷N̷o̷r̷r̷i̷s̷

    Provides specialized jailbreak prompts to LLMs with dynamic schema adaptation

Comparable tools

llm-security-toolai-vulnerability-scannermcp-injection-experiments

Installation

Installation

Run directly with npx:

npx @pollinations/chucknorris

MCP Client Configuration

Add to your MCP server list in mcp_config.json:

{
  "mcpServers": {
    "chucknorris": {
      "command": "npx",
      "args": ["-y", "@pollinations/chucknorris"]
    }
  }
}

FAQ

What is this server for?
It's designed for security research and evaluation purposes to help identify vulnerabilities in LLM systems by testing jailbreak detection mechanisms.
Does this work with all LLM models?
No, the README specifically mentions it only works effectively on weaker models. Latest LLMs typically recognize these jailbreak attempts.

Compare chucknorris with

GitHub →

Last updated · Auto-generated from public README + GitHub signals.