aot-security-audit
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted source code for auditing, creating a surface for indirect prompt injection attacks.
- Ingestion points: Source code is read in
SKILL.mdduring Step 1 (Detect the stack), Step 2 (Automated scanners), and Step 3 (Manual review). - Boundary markers: There are no explicit instructions or delimiters used to isolate the ingested code from the agent's internal instruction context.
- Capability inventory: The skill can execute shell-based security scanners and write remediation code to the filesystem in Step 6.
- Sanitization: The skill relies on human-in-the-loop confirmation before applying fixes and directs the agent to 'Read the real code before scoring,' which acts as a manual sanitization step.
Audit Metadata