The Agent Skills Directory

[COMMAND_EXECUTION]: Uses local shell commands including git grep, git ls-files, npm audit, and pip-audit to inspect repository health and identify vulnerabilities.
[DATA_EXFILTRATION]: Identifies sensitive file patterns such as .env, .pem, and hardcoded credentials; however, it lacks any network-bound commands (curl, wget, fetch) that would exfiltrate this discovered data.
[SAFE]: The skill's primary purpose is security auditing. The logic is transparently documented, focuses on identification of risks, and adheres to a read-only workflow as stated in its prerequisites.

evaluate-repository