Strix

This skill wraps Strix, an AI-driven security testing CLI that runs vulnerability scans inside a Docker sandbox using an LLM for reasoning. You point it at local code, GitHub repos, or live URLs you're authorized to test, and it hunts for security issues using configurable scan depths from quick PR checks to deep bug bounty style exploration. The setup is straightforward if you've got Docker running and an OpenAI-compatible API key. It handles both white-box source analysis and black-box live testing, can combine multiple targets in one scan, and returns structured results with exit codes you can wire into CI. The instruction file approach for passing credentials and scope rules is cleaner than most pentest automation tools I've seen.

Install to Claude Code

npx -y skills add akillness/oh-my-skills --skill strix --agent claude-code