The Agent Skills Directory

[SAFE]: The skill implements a human-in-the-loop safety pattern by requiring the agent to present a 'kill list' and wait for user review before proceeding with any file deletions.
[SAFE]: The logic provided for culling tests (e.g., existence tests, typecheck tests) is centered on software testing best practices and does not include any commands for privilege escalation, network exfiltration, or persistence.
[SAFE]: No obfuscation, hardcoded credentials, or remote code execution patterns were detected in the instructions.

cull-tests