The Agent Skills Directory

[SAFE]: The skill implements a structured interview persona with specific constraints to ensure quality and focus, containing no attempts to bypass safety or ethical guidelines.
[SAFE]: Command execution is limited to a benign directory existence check (ls -d .knowhow/) to confirm project initialization before proceeding.
[SAFE]: File access is restricted to reading local logs and user-provided transcripts for context, which is standard for knowledge management and does not involve sensitive system paths.
[SAFE]: No indicators of data exfiltration, obfuscation, or remote code execution were found in the instructions or metadata.

knowhow-reflect