The Agent Skills Directory

[SAFE]: The skill implements a persona for critical interrogation and stress-testing. While the instructions use terms like 'hostile' and 'rude', they are explicitly constrained to attacking ideas and claims rather than the user, providing a structured boundary for the adversarial framing.
[SAFE]: The skill maintains session state by reading and writing to a local markdown file in the working directory. This is a standard practice for tracking progress and does not involve sensitive directories or system files.
[SAFE]: No external network dependencies, remote code execution patterns, or credential exposures were found. All resources and concepts referenced are consistent with the skill's stated purpose.

asb-rude-qa