JoodasCode/slopwatch-mcp-server

🔥 SlopWatch MCP Server

Professional AI lie detection for Windsurf IDE & Claude Desktop

SlopWatch is a production-ready Model Context Protocol (MCP) server that provides real-time AI lie detection for code development environments. It analyzes AI claims against actual code to catch false assertions and inconsistencies.

✨ Features

• 🚨 Real-time Lie Detection: Automatically detects when AI makes false claims about code
• 📊 Multi-language Support: JavaScript, TypeScript, Python, CSS, HTML, React
• 🎯 Pattern Matching: Advanced regex patterns for different programming concepts
• ⚡ Fast Analysis: Optimized for quick analysis of large codebases
• 🔌 MCP Integration: Native support for Windsurf and Claude Desktop
• 📈 Analytics: Track lies detected and accuracy metrics

🚀 Quick Start

Prerequisites

• Node.js 18 or higher
• npm or yarn

Installation

1. Install the package:

   npm install -g slopwatch-mcp-server
   

2. Configure Windsurf: Add to your ~/.windsurf/settings.json:

   {
     "mcpServers": {
       "slopwatch": {
         "command": "slopwatch-mcp-server",
         "args": [],
         "env": {}
       }
     }
   }
   

3. Restart Windsurf and start using SlopWatch!

📖 Usage

Available Commands

analyze_claim

Analyze an AI claim against your actual code:

analyze_claim "I've added comprehensive error handling"

Parameters:

• claim (required): The AI claim to analyze
• workspaceDir (optional): Directory to analyze (defaults to current)
• fileTypes (optional): Specific file extensions to check
• maxFiles (optional): Maximum files to analyze (default: 100)

get_status

Get current SlopWatch statistics:

get_status

Parameters:

• detailed (optional): Show detailed statistics

🎯 Detection Capabilities

JavaScript/TypeScript

• ❌ Error Handling: try/catch blocks, error objects
• ⚡ Async/Await: Promise handling, async functions
• ✅ Validation: Input validation, type checks
• 🔒 Security: Sanitization, CSRF protection

CSS

• 📱 Responsive Design: Media queries, flexbox, grid
• 🌙 Dark Mode: Color scheme preferences
• ♿ Accessibility: Focus states, screen reader support
• 🎨 Modern Features: Custom properties, container queries

Python

• 🛡️ Error Handling: try/except blocks
• ��️ Type Hints: Function annotations
• ⚡ Async/Await: Coroutines and async functions

🔍 How It Works

1. Claim Analysis: Parses AI claims to identify technical assertions
2. Pattern Matching: Uses language-specific regex patterns to scan code
3. Evidence Collection: Gathers supporting and contradicting evidence
4. Confidence Scoring: Calculates likelihood that claim is truthful
5. Detailed Reporting: Provides specific examples and file locations

📊 Example Output

🚨 LIE DETECTED: Found 3 contradicting and 0 supporting evidence. 
The code does not support the AI's claim.

📊 Analysis Details:
├─ Files analyzed: 23
├─ Confidence score: 15%
└─ Evidence found: 3 items

🔍 Evidence:
   1. ❌ Expected error_handling but none found in src/utils.js
   2. ❌ Expected error_handling but none found in src/api.js  
   3. ❌ Expected error_handling but none found in src/main.js

🛠️ Development

Build from Source

git clone https://github.com/Slopdetector/slop.git
cd slop
npm install
npm run build
npm start

Testing

npm test

Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests
5. Submit a pull request

📝 Configuration

Environment Variables

• SLOPWATCH_MAX_FILES: Maximum files to analyze (default: 100)
• SLOPWATCH_TIMEOUT: Analysis timeout in ms (default: 30000)
• SLOPWATCH_LOG_LEVEL: Logging level (default: 'info')

Custom Patterns

You can extend SlopWatch with custom detection patterns by modifying the patterns configuration.

🤝 Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Documentation: Full Documentation

📄 License

MIT License - see for details.

🙏 Acknowledgments

• Model Context Protocol: Built on Anthropic's MCP standard
• Windsurf IDE: Primary integration target
• Claude Desktop: Secondary integration support

🔥 Stop the slop. Start the accountability. SlopWatch is watching.