safe-to-run

Safe to Run

Why This Exists

AI agents are now deeply involved in software development, and code generation velocity has gone up tremendously. The volume of new repos, libraries, and tools being published is hard to keep up with — and star, fork, contributor counts no longer mean what they used to.

Even a polished repo with thousands of stars can have bad intentions baked in: a malicious postinstall script, a dependency that was quietly hijacked, or telemetry that phones home on startup.

I run this skill before installing anything open source, and I recommend you do the same. A quick audit takes seconds and can catch something you'd never notice by skimming a README.

Install

npx skills add bgultekin/safe-to-run

Skills

• safe-to-run: Performs a thorough security audit of a code repository — checks for malicious intent, supply chain risks, unintentional harm, and suspicious network behavior. Produces a structured findings report with a final ✅ / ⚠️ / 🚫 verdict.

License

MIT