user-persona
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read and process user-provided files such as research data, interview transcripts, survey results, and analytics, which may introduce an indirect prompt injection surface if those external inputs contain adversarial instructions.
- Ingestion Points:
SKILL.md(Context section specifies processing user-provided documents and data files). - Boundary Markers: Absent. The instructions do not define strict boundaries or delimiters around the ingested text to separate content from instructions.
- Capability Inventory:
SKILL.md(Instructs the agent to save output as a markdown document in the user's workspace). - Sanitization: Absent. No sanitization or safety filtering is specified for the input data text before processing.
Audit Metadata