user-segmentation-profiler
Pass
Audited by Gen Agent Trust Hub on Jul 2, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by processing untrusted user input and storing it for use by other skills.
- Ingestion points: Conversation history and direct user answers about experience and goals in SKILL.md.
- Boundary markers: No specific delimiters or warnings to ignore instructions inside user data are present.
- Capability inventory: The skill writes processed data to
memory/user_profile.md. - Sanitization: There is no evidence of input validation or sanitization before writing to the profile file.
Audit Metadata