review-audit
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted customer reviews from external sources, creating a potential surface for indirect prompt injection.
- Ingestion points: Customer reviews are ingested via chat, CSV, or spreadsheet uploads as described in the SKILL.md file.
- Boundary markers: The instructions do not define explicit delimiters or instructions for the agent to ignore potentially malicious commands embedded within the reviews.
- Capability inventory: No high-risk capabilities, such as network operations, file system modifications, or shell command execution, were identified in the skill instructions.
- Sanitization: The instructions do not specify any input validation, filtering, or sanitization routines for the ingested review content.
Audit Metadata