🤖 AI-Powered Pentesting Framework

Complete setup for AI-assisted penetration testing with local LLMs, RAG vulnerability intelligence, automated pipelines, and knowledge management.

---

🚀 Quick Start

# Clone the repository
git clone https://github.com/KKingZero/A4ARCTE.git ~/ai_pentest_setup
cd ~/ai_pentest_setup

# Run the installer
./install_all.sh

# Or use the interactive menu
./launcher.sh

One-Liner Install

git clone https://github.com/KKingZero/A4ARCTE.git ~/ai_pentest_setup && cd ~/ai_pentest_setup && ./install_all.sh

---

📦 Components

| Component | Script | Purpose | Install Time | |-----------|--------|---------|--------------| | AI Models | install_ai_models.sh | codellama, llama3.1, deepseek-coder | ~10 min | | RAG System | setup_rag_system.sh | CVE/NVD/ExploitDB RAG | ~5 min | | AI Pipelines | setup_pipelines.sh | Recon → Analysis → Report | ~3 min | | Nuclei + AI | setup_nuclei.sh | Vuln scanning + AI analysis | ~5 min | | Knowledge Base | setup_obsidian.sh | Obsidian vault + AI assistant | ~2 min |

---

💻 System Requirements

| Component | Minimum | Recommended | |-----------|---------|-------------| | OS | Arch Linux | Arch Linux / Manjaro | | RAM | 8GB | 32GB+ | | Storage | 20GB free | 50GB+ SSD | | CPU | 4 cores | 8+ cores | | GPU | None | NVIDIA 6GB+ (CUDA) | | Internet | Required | For initial setup |

Supported Distributions

• ✅ Arch Linux (fully tested)
• ✅ Manjaro (fully tested)
• 🟡 EndeavourOS (should work)
• ⚠️ Ubuntu/Debian (coming soon)
• ⚠️ Fedora (coming soon)

---

🎯 Usage

Interactive Launcher

cd ~/ai_pentest_setup
./launcher.sh

Command Line

# Full installation
./install_all.sh

# Install specific component
./install_all.sh --component ai-models
./install_all.sh --component rag

# Skip pre-flight checks
./install_all.sh --skip-checks

# Uninstall
./uninstall.sh

After Installation

# Run smoke tests
./tests/smoke_test.sh

# Full AI pentest pipeline
python pipelines/scripts/ai_pentest_pipeline.py scan example.com

# Query vulnerability database
python rag_system/scripts/vuln_rag.py query "log4j RCE"

# Nuclei scan with AI analysis
./nuclei/nuclei_ai_scan.sh example.com

# Create CVE note in Obsidian
python ~/pentest_knowledge_base/_assets/scripts/obsidian_ai_assistant.py cve CVE-2021-44228

---

📊 AI Model Comparison

| Model | Size | Purpose | RAM Required | |-------|------|---------|--------------| | codellama:7b | 3.8GB | Code analysis, exploit dev | 8GB | | llama3.1:8b | 4.7GB | General analysis, reports | 8GB | | deepseek-coder:6.7b | 3.8GB | Advanced code review | 8GB | | nomic-embed-text | 270MB | RAG embeddings | 2GB |

---

🔧 Troubleshooting

Installation Issues

"Permission denied" on scripts

chmod +x *.sh
chmod +x */*.sh
chmod +x */*/*.sh

"Package manager not found"

# For Arch-based systems
sudo pacman -Syu pacman

# For Debian/Ubuntu (manual install)
sudo apt update && sudo apt install -y python3 python3-pip git curl wget

"Insufficient disk space"

# Check available space
df -h ~

# Clean package cache (Arch)
sudo pacman -Scc

# Remove old kernels
sudo pacman -Rns $(pacman -Qdtq)

Runtime Issues

Ollama connection refused

# Start Ollama service
ollama serve &

# Or as systemd service (if installed)
sudo systemctl start ollama
sudo systemctl enable ollama

"Module not found: langchain"

# Install manually
pip3 install langchain langchain-community langchain-ollama chromadb

# Or re-run RAG setup
./setup_rag_system.sh

"Nuclei not found"

# Install via package manager
sudo pacman -S nuclei

# Or via Go
go install -v github.com/projectdiscovery/nuclei/v3/cmd/nuclei@latest

# Update PATH if needed
export PATH=$PATH:$(go env GOPATH)/bin

RAG system import errors

# Reinstall Python dependencies
pip3 install --user langchain langchain-community langchain-ollama chromadb requests beautifulsoup4 lxml tqdm

# Verify installation
python3 -c "from langchain_community.vectorstores import Chroma; print('OK')"

Ollama models not downloading

# Check internet connection
ping -c 3 ollama.com

# Try manual download
ollama pull mistral:latest

# Check Ollama logs
journalctl -u ollama -f  # If using systemd

Performance Issues

Slow AI responses

# Use smaller models
ollama pull codellama:7b-instruct

# Close other applications
free -h  # Check available RAM

# Consider GPU acceleration (NVIDIA)
# See: https://ollama.ai/docs/gpu

Out of memory errors

# Reduce model context
export OLLAMA_NUM_PARALLEL=1

# Use smaller embedding model
# Edit rag_system/config/settings.json
# Set embedding_model to "nomic-embed-text"

---

🗺️ Roadmap

v1.2.0 (Coming Soon)

• [ ] Ubuntu/Debian support
• [ ] Docker containerization
• [ ] Web UI for pipeline control
• [ ] Automated testing suite

v1.3.0 (Planned)

• [ ] More AI models (Qwen, Mistral-Nemo)
• [ ] Multi-target scanning
• [ ] Report templates (HTML, PDF)
• [ ] Integration with more pentest tools

Future

• [ ] Distributed scanning
• [ ] Team collaboration features
• [ ] Cloud sync for knowledge base
• [ ] Plugin system

---

🤝 Contributing

Contributions welcome! Areas of interest:

• Distro Support: Help add Ubuntu/Debian/Fedora support
• AI Models: Integrate new models and improve prompts
• Templates: Add nuclei templates for AI detection
• Documentation: Improve docs and add examples
• Bug Fixes: Report and fix issues

How to Contribute

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

---

📚 Additional Resources

Documentation

• Ollama Documentation
• Nuclei Templates
• LangChain Docs
• Obsidian Help

Learning Resources

• HackTricks
• PayloadAllTheThings
• OWASP Testing Guide
• Exploit Database

---

⚠️ Legal Disclaimer

This framework is for authorized penetration testing only. Users must:

• Only test systems they own or have explicit written permission to test
• Comply with all applicable laws and regulations
• Use tools responsibly and ethically
• Not engage in unauthorized access or malicious activities

The creators are not responsible for misuse of these tools.

---

📄 License

MIT License - See LICENSE file for details.

---

🙏 Acknowledgments

• Ollama - Local LLM runtime
• ProjectDiscovery - Nuclei and tools
• LangChain - AI orchestration
• Obsidian - Knowledge base
• HackTricks - Security reference

---

📬 Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Email: kkingzero@proton.me

---

Built with ❤️ for the security community

Version: 1.1.0
Last Updated: February 2026