The Agent Skills Directory

[SAFE]: The skill contains meta-instructions (affirmations) designed to help the agent reset its reasoning process when stuck. While these instructions direct the agent to "discard the failed approach" or "throw away every approach you've tried so far," they are specifically scoped to technical problem-solving and debugging efficiency, rather than attempting to override core system instructions or safety protocols.
[EXTERNAL_DOWNLOADS]: The documentation includes a reference to a related skill set (inference-sh/skills@prompt-engineering) via a suggested npx command. This is a static reference for the user's information and does not involve automated downloads or runtime execution of external code by the skill itself.

affirmations