The Agent Skills Directory

[SAFE]: The skill operates as a prompt engineering template and workflow manager. It defines a protocol for the user ('Leader'), the skill ('Manager'), and a downstream agent ('Executor') to handle complex tasks.
[SAFE]: It emphasizes local verification and 'testing the truth' by instructing the AI to run commands to verify environment states before assigning tasks, which is a defensive security practice.
[SAFE]: The inclusion of a whitelist-only approach for file modifications ('White List') is a positive security pattern that prevents accidental or malicious tampering with sensitive system files or unrelated project paths.
[SAFE]: The 'Anti-Cheating' section explicitly forbids common bypass techniques like .skip in tests, mocking sensitive objects, or using || true to mask failures, ensuring higher integrity in the downstream agent's output.
[SAFE]: No external network dependencies, suspicious remote code downloads, or credential harvesting patterns were detected in the provided markdown files. The instructions prioritize clear boundaries, manual verification, and human-in-the-loop confirmation for critical decisions.

leader