The Agent Skills Directory

[SAFE]: No security issues were detected. The skill performs audits on change management records using a structured multi-agent workflow and local shell commands (mkdir, date, cat) for 'saga' state management and time-based break-circuits. All file operations are restricted to the local .aidoc directory. While the skill processes untrusted document content, it proactively employs security best practices by explicitly labeling artifact data as 'untrusted' and instructing the agent to ignore specific metadata fields (de-anchoring) to prevent bias and prompt injection risks.

doc-chg-audit