interactive-requirements-gathering
Pass
Audited by Gen Agent Trust Hub on Mar 6, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by processing untrusted user data.\n- Ingestion points: User responses to questionnaires, particularly the free-form 'Type your own answer' option (Option D) defined in SKILL.md.\n- Boundary markers: The framework lacks explicit delimiters or 'ignore' instructions to prevent the agent from potentially following malicious commands embedded within user answers.\n- Capability inventory: The skill utilizes the Read, Write, and Edit tools to create or modify files based on gathered requirements, creating an exploitable path for injected content.\n- Sanitization: No sanitization or validation logic is defined to filter instructions from user-provided text before it is used in content generation.
Audit Metadata