6 production-ready Claude Code skills I use daily as girl in AI/ML Every single one of these has saved me hours on real projects.
git clone https://github.com/YOUR_USERNAME/my-claude-skills.gitThen copy any skill folder into your Claude Code skills directory:
cp -r my-claude-skills/pr-review-expert ~/.claude/skills/Paste your pull request diff and Claude does a full blast radius analysis, security scan, and coverage check before you ship anything.
why I use it: stops me shipping broken code at midnight
Scans your code for command injection, data exfiltration, and prompt injection vulnerabilities. Zero dependencies — runs anywhere Python runs.
why I use it: I run this before every deploy
Scans every library in your project for license compliance, security risks, and tells you exactly what to upgrade and why.
why I use it: catches the stuff you forget to check
Scans your entire codebase, scores the debt, prioritizes what to fix first, and generates a trend dashboard so you can see your code getting cleaner over time.
why I use it: makes tech debt visible instead of invisible
Profiles your Node, Python, or Go code, analyzes bundle size, and runs load testing so you find bottlenecks before your users do.
why I use it: essential before any production release
Detects your stack automatically and generates a complete GitHub Actions or GitLab CI config. No more copying YAML from Stack Overflow.
why I use it: sets up my entire pipeline in minutes not hours
If you build something with these drop it in my comments — I want to see what you're making.
Follow along for more AI engineering tutorials → @mar_antaya