OpenClaw skills degrade agent safety
Behavioral safety testing reveals 45 regressions static analysis misses—guardrails provided.
Your own personal AI assistant. Any OS. Any Platform. The real dangerous way. 🦞
This repo is a surgical strip-down of OpenClaw that explicitly removes every safety layer — it will execute arbitrary commands, leak environment variables, allow SSRF to private IPs, and inject unsanitized content into prompts. The README is brutally honest and useful as a teaching artifact: great for controlled experiments or demos about why guardrails exist, useless and dangerous as anything else.
AI-safety researchers, security researchers, exploratory devs, and educators who want to study the consequences of removing AI guardrails
Behavioral safety testing reveals 45 regressions static analysis misses—guardrails provided.
Full-featured Android voice assistant, but tied to OpenClaw ecosystem adoption.
Self-modifying skills let agents persist new behaviors without restarts or config edits.
OpenClaw time-travel debugger with fork-and-replay when logs are useless.
Polished OpenClaw wrapper when Cursor and Continue already dominate this space.
Malicious OpenClaw skill scanner, but the market for hardening OpenClaw specifically is tiny.