The Agent Skills Directory

[SAFE]: The skill operates as a linguistic and structural framework for critiquing artifacts. It contains no functional code, external network calls, or file system access.
[SAFE]: The instructions include an 'Auto-clarity exception' in SKILL.md, which explicitly directs the agent to suspend the skill's stylistic constraints when performing high-stakes, destructive, or security-related operations (such as rm -rf or security warnings). This ensures that critical system feedback remains clear and unhindered.
[SAFE]: No obfuscation, data exfiltration patterns, or persistence mechanisms were detected. The skill's 'anchor' mode is a standard instructional technique for setting a specific persona or behavioral guideline within a session.

taste