The Agent Skills Directory

[SAFE]: The skill is a purely instructional module that provides a framework for performing security audits. There are no patterns of prompt injection, privilege escalation, or persistence mechanisms detected within its content.
[NO_CODE]: This skill does not contain any executable scripts, binaries, or external dependencies. It is comprised entirely of a markdown file with system instructions, which significantly reduces its attack surface.
[DATA_EXFILTRATION]: While the skill requests 'fileRead' access to analyze other skill files, it does not request 'network' permissions. This configuration ensures that any information read by the agent cannot be sent to external servers, adhering to the principle of least privilege.

permission-auditor