The Agent Skills Directory

[COMMAND_EXECUTION]: The script utilizes subprocess.run to call internal platform-runtime components for session transcript processing. This is implemented securely using a list of arguments rather than a shell string, ensuring that inputs such as the session ID are not executed as commands.
[SAFE]: The skill incorporates robust file system protections to prevent path traversal. Specifically, the print-phase-breakdown operation validates that user-supplied output paths are relative and remain within the authorized plan directory using resolved path checks.

manage-metrics