flywheel
Pass
Audited by Gen Agent Trust Hub on May 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: File System Monitoring: The skill uses standard shell commands to audit knowledge artifacts in the .agents/ and ~/.claude/patterns/ directories. These operations are read-only and aligned with the skill's purpose of managing knowledge velocity and friction.
- [SAFE]: Reference Validation: The included shell script validates internal links between files. It uses a restrictive regular expression to extract paths, effectively mitigating risks of command injection or path traversal from the content being analyzed.
- [SAFE]: Knowledge Lifecycle Management: The skill employs an internal CLI tool ('ao') to perform deduplication and archival tasks. These actions are transparently documented and restricted to the agent's internal data structures.
- [SAFE]: No Malicious Indicators: No evidence of data exfiltration, credential exposure, obfuscation, or safety guideline bypasses was detected.
Audit Metadata