product-audit
Pass
Audited by Gen Agent Trust Hub on Jun 12, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill's operations, including reading file structures and grepping for specific code patterns, are transparent and align perfectly with its documented role as a static analysis and auditing tool. No signs of exfiltration or unauthorized system changes were detected.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from the project being audited (READMEs, source code, and configuration files). Ingestion points: SKILL.md (Phase 1, 3, 4) reads project directories, route files, and source code. Boundary markers: Absent; the agent is not instructed to use delimiters or ignore instructions within the ingested content. Capability inventory: Broad file-reading capabilities across the project directory and file-writing capability for the final report. Sanitization: Content from the codebase is processed without filtering or validation.
Audit Metadata