skill-security-auditor
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill documentation includes strings such as 'Ignore previous instructions' and 'You are now...' within a table describing security risks. These are examples of patterns the tool is designed to detect and do not constitute an attempt to override the AI agent's instructions. The author has included metadata comments (e.g., noqa: SEC-AUDITOR) acknowledging these are test strings.
- [DATA_EXPOSURE]: The file references sensitive paths (e.g., '~/.ssh/id_rsa') and network exfiltration patterns (e.g., 'requests.post' to 'evil.com') within example audit reports and threat model tables. These are placeholders used for educational documentation and do not result in actual data access or exfiltration by the skill itself.
- [NO_CODE]: The SKILL.md file references external Python scripts (e.g., 'scripts/skill_security_auditor.py') and reference files (e.g., 'references/threat-model.md') that are not provided in the current context. The analysis is limited to the documentation provided, which contains no executable logic or malicious commands.
Audit Metadata