product-manager
Fail
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: CRITICALPROMPT_INJECTIONNO_CODE
Full Analysis
- [Indirect Prompt Injection] (LOW): The skill processes untrusted user requirements to generate documentation, presenting an indirect prompt injection surface. 1. Ingestion points: User requirements input (SKILL.md Step 1). 2. Boundary markers: Absent. 3. Capability inventory: File-write operations to the outputs/ directory (SKILL.md Step 5). 4. Sanitization: Absent.
- [External Downloads] (SAFE): The automated scan flagged 'requirements.md' as a malicious URL, but this is a false positive triggered by a local filename. No actual remote downloads or execution patterns were identified.
- [Data Exposure] (SAFE): No hardcoded credentials or access to sensitive system files (e.g., SSH keys, AWS configs) were detected within the skill files or templates.
Recommendations
- Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata