The Agent Skills Directory

[SAFE]: The skill provides stylistic and operational guidelines for the AI agent (e.g., 'telegram-terse replies', 'earn every test'). These are standard behavioral constraints and do not represent security risks.
[SAFE]: Instructions regarding shared machine usage (like pkill or database resets) explicitly mandate that the agent must 'ask first', which is a safety-oriented instruction rather than an attempt to perform unauthorized actions.
[SAFE]: References to other skills (ponytail, conventional-commits) are internal tool-use instructions and do not involve remote code execution or external downloads.

standing-orders