AI • CLI • Security
VulnClaw
Product Delivery: AI-Powered CLI from Zero to Open Source Community
Built an open-source AI penetration testing CLI from scratch. Managed the full product lifecycle: scope definition, architecture design, community building, roadmap prioritization, and shipping v0.4.0 with active users across 3 continents. The tool integrates AI agents for autonomous reconnaissance, vulnerability scanning, and report generation. Key lesson: starting narrow beats building broad every time. 1,200+ GitHub stars, 40+ forks, 15+ community contributors.
Features
Capabilities
Product scoping: evolved from a simple port scanner to a multi-agent AI penetration testing framework in 6 months (v0.1 → v0.4)
AI agent architecture: 3 specialized agents — Recon, Exploit, Report — that collaborate autonomously using LLM reasoning
Community management: triaged 80+ PRs and 200+ issues, maintaining a 48h first-response SLA for critical security reports
Roadmap prioritization: said 'no' to 60% of feature requests to ship the essential 40% — ruthless scope control
Open source as delivery strategy: MIT license, GitHub Discussions for roadmap voting, automated CI/CD with GitHub Actions
Tech Stack
Python
Typer
Rich
Prompt Toolkit
AI
Screenshots
Terminal-first experience
CLI help & commands
Live reconnaissance scan
Findings & results
Get VulnClaw
VulnClaw is proprietary. Contact me for licensing inquiries or check out the project on GitHub.