The Agent Skills Directory

[COMMAND_EXECUTION]: The file SKILL.md contains instructions for the agent to execute shell commands such as git check-ignore and to modify the .git/info/exclude file. These commands are intended for local workspace management, specifically to prevent temporary draft notes generated during verification from being committed to the version control system.
[SAFE]: The core logic of the skill is centered on improving the quality of AI responses. It establishes a hierarchy of evidence and sources, encouraging the agent to cross-reference technical details with live documentation and code rather than relying solely on potentially outdated training data. No malicious intent, exfiltration patterns, or obfuscation techniques were detected.

best-practices