customer-persona-framer
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external untrusted data sources, including customer research, interview notes, and observations, as specified in both SKILL.md and references/journey-framing.md. This ingestion point creates a potential surface for indirect prompt injection if the input materials contain hidden or adversarial instructions. While the instructions include a 'Data Sufficiency Gate' to logically separate research findings from inferences, the skill lacks technical sanitization, explicit boundary markers, or specific instructions for the agent to ignore potentially malicious embedded content within the source data.
Audit Metadata