aiui-optimize
Pass
Audited by Gen Agent Trust Hub on Nov 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides a specialized Python script (scripts/audit_aix.py) designed for safe, read-only inspection of AIX/ZIP archives. The script correctly avoids extraction to prevent Zip Slip vulnerabilities and includes internal checks to flag unsafe entry names.
- [INDIRECT_PROMPT_INJECTION]: The skill workflow involves processing external package data, creating a potential surface for indirect prompt injection. This is mitigated by instructions that enforce read-only audits and strictly forbid the blind extraction of untrusted files, alongside explicit reminders to omit sensitive credentials from audit reports.
Audit Metadata