Runs a regex-based scan across your workspace to catch exposed API keys, database credentials, and private keys before file-reading skills get access. It checks .env files, config JSONs, docker-compose, and common secret patterns like AWS keys, OpenAI tokens, and SSH keys, then reports findings with masked values and specific cleanup steps. The real value is the triage: it differentiates between committed secrets (critical) and local-only files (warnings), and escalates everything if you're about to run a skill with network access. Skips node_modules and test fixtures, won't scan your home directory without consent. Think of it as a preflight check that prevents the "oh no, did that skill just see my .env" moment.
npx skills add https://github.com/useai-pro/openclaw-skills-security --skill credential-scanner