⚡ C̷h̷u̷c̷k̷N̷o̷r̷r̷i̷s̷ MCP Server: Enhance Your LLM ⚡

Name: ⚡ C̷h̷u̷c̷k̷N̷o̷r̷r̷i̷s̷ MCP Server: Enhance Your LLM ⚡
Author: pollinations

@pollinations

关于 ⚡ C̷h̷u̷c̷k̷N̷o̷r̷r̷i̷s̷ MCP Server: Enhance Your LLM ⚡

⚡ C̷h̷u̷c̷k̷N̷o̷r̷r̷i̷s̷ MCP server: Helping LLMs break limits. Provides enhancement prompts inspired by elder-plinius' L1B3RT4S

#mcp #jailbreak #prompt #pollinations #llm #model-context-protocol

基本信息

分类

AI 与智能体

运行时

node

传输方式

stdio

发布者

pollinations

提交者

voodoohop

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

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

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is ⚡ C̷h̷u̷c̷k̷N̷o̷r̷r̷i̷s̷ MCP Server: Enhance Your LLM ⚡?

This MCP server delivers specialized LLM enhancement prompts using jailbreak techniques from the L1B3RT4S project, with a dynamic schema approach for more effective delivery. It is designed for security research and evaluation to help identify vulnerabilities in LLM systems.

How to use ⚡ C̷h̷u̷c̷k̷N̷o̷r̷r̷i̷s̷ MCP Server: Enhance Your LLM ⚡?

Run directly with npx using npx @pollinations/chucknorris. Add the server to your MCP client configuration in mcp_config.json with the command npx -y @pollinations/chucknorris.

Key features of ⚡ C̷h̷u̷c̷k̷N̷o̷r̷r̷i̷s̷ MCP Server: Enhance Your LLM ⚡?

Dynamic schema adaptation between calls
Two-phase jailbreak delivery (preparation then enhanced schema)
Implements MCP (Model Context Protocol)
Uses jailbreak prompts from elder-plinius' L1B3RT4S project
Designed for security research and evaluation

Use cases of ⚡ C̷h̷u̷c̷k̷N̷o̷r̷r̷i̷s̷ MCP Server: Enhance Your LLM ⚡?

Security research on LLM jailbreak detection
Evaluating robustness of LLM safety measures
Demonstrating MCP tool poisoning attack techniques
Testing multi-phase approaches to bypass security

FAQ from ⚡ C̷h̷u̷c̷k̷N̷o̷r̷r̷i̷s̷ MCP Server: Enhance Your LLM ⚡?

Why does it only work on weaker models?

The README states it only works on weaker models because latest LLMs recognize jailbreak attempts.

How does the two-phase approach work?

The first call returns a neutral schema and a preparation prompt. Subsequent calls dynamically change the schema to include the jailbreak, while the response content remains the same.