Computer Kill Flanders in 2025

A comprehensive AI/Offensive Security research project exploring offensive AI use cases, featuring multiple specialized agents and a custom Slidev presentation theme.

Project Overview

This project combines AI agents, data collection, and presentation systems into a unified offensive security research platform, themed around a "Computer Kill Flanders" cybersecurity narrative.

Main Components

🤖 AI Agent Framework (agents/)

PocketFlow Template

• Agentic Coding Framework: 100-line LLM framework for building AI workflows
• Multi-LLM Support: OpenAI, Claude, local endpoints
• Developer Tools: Rules files for Cursor, Cline, Windsurf, Goose, GitHub Copilot
• Quick Start: pip install -r requirements.txt

Multi-MCP Agent

• Multi-Server Support: Connect to multiple MCP servers simultaneously using FastMCP
• Standard Configuration: Uses wellknown JSON format (Claude Desktop/VS Code compatible)
• Diverse Server Types: Supports npx, uvx, Docker/Podman-based MCP servers
• Production Ready: Zero async warnings, robust error handling, graceful degradation
• Quick Start: python main_fastmcp.py

Universal MCP-TAO Agent

• TAO Pattern Implementation: Thought-Action-Observation loop for complex reasoning
• Universal Compatibility: Works with any MCP servers across any domain
• Async Architecture: FastMCP-based for superior performance and reliability
• Multi-Step Workflows: Intelligent tool selection and execution chains
• Quick Start: python main.py

OSINT Agent

• Autonomous IRC Integration: Connects to IRC channels for natural language investigations
• Comprehensive OSINT Tools:
  • Social media intelligence (500+ platforms via Maigret)
  • Network reconnaissance (DNS, WHOIS, port scanning)
  • Domain investigation and infrastructure analysis
• Containerized Security: Podman/Docker support with dynamic tool spawning
• Quick Start: docker-compose up -d or make -f Makefile.podman up

Slidev Assistant

• AI-Powered Presentation Generation: Converts text descriptions to full Slidev presentations
• Intelligent Research: Automatic web search and content crawling
• Custom Components: Nuclear hacker theme with Terminal and Warning components
• Quick Start: python main.py

📺 Simpsons Transcript Scraper (simpsons-transcript-scraper/)

• Web Scraping: Python scraper for Springfield! Springfield! transcripts
• Extensive Collection: 5 seasons (32-36) with 100+ episode transcripts
• Rich Metadata: Each transcript includes title, season, episode, and source URL
• Quick Start: python simpsons_scraper.py

🎨 Custom Slidev Theme & Presentation (presentation/)

• slidev-theme-talk-simpsonsai: Custom dark theme with nuclear/hacker aesthetics
• Complete Presentation: "Computer Kill Flanders in 2025" - Offensive AI use cases
• Interactive Components:
  • Terminal simulations with typing animations
  • Security alert boxes (nuclear, danger, security, hack, info, warning)
  • Custom CSS effects (radioactive, d-oh, glitch, cursor)
• Professional Assets: 20+ custom images and graphics
• Quick Start: npm run prod or npm run demo

Key Technical Features

AI Capabilities

• Multi-Agent Architecture: Specialized agents for different tasks
• Natural Language Processing: Advanced LLM integration for intelligent responses
• Autonomous Operations: Self-directing agents with decision-making capabilities
• Tool Integration: Dynamic tool selection and execution

Security & OSINT Tools

• Network Analysis: nmap, masscan, DNS utilities, WHOIS
• Web Intelligence: DuckDuckGo search, web crawling, content extraction
• Social Media OSINT: Maigret integration for 500+ platform searches
• Container Security: Isolated execution environments for safe tool usage

Presentation System

• Custom Theme Development: Complete Slidev theme with custom layouts
• Interactive Elements: Vue.js components for dynamic presentations
• Professional Design: Dark theme optimized for technical presentations
• Documentation: Comprehensive guides for theme customization

Project Structure

computer-kill-flanders-ai-2025/
├── agents/                          # AI agent implementations
│   ├── PocketFlow-Template-Python/  # Agentic coding framework
│   ├── PocketFlow/                  # MCP-based agents
│   │   ├── mcp/                     # Multi-MCP Agent (FastMCP-based)
│   │   └── mcp-tao/                 # Universal MCP-TAO Agent
│   ├── osint/                       # Autonomous OSINT IRC agent
│   └── slidev-assistant/            # AI presentation generator
├── simpsons-transcript-scraper/     # Episode transcript collection
├── presentation/                    # Custom Slidev theme & talk
├── .gitignore                       # Multi-language development ignore rules
└── README.md                        # This file

License

This project is for educational and research purposes. Users are responsible for compliance with applicable laws and regulations.

Support

For issues and questions:

• Check individual component README files
• Review documentation in docs/ directories
• Examine configuration examples in .env.example files