lyt-detail-page
Pass
Audited by Gen Agent Trust Hub on Jul 29, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface through the ingestion of untrusted external content like competitor pages and reference images.
- Ingestion points: User-uploaded competitor detail pages and visual reference materials (SKILL.md).
- Boundary markers: Specific 'Facts Gate' and 'Slogan Rules' serve as instructions to isolate and ignore unverified claims from external sources.
- Capability inventory: Capabilities are restricted to text generation and the use of image generation tools, with no access to subprocesses, shell commands, or network exfiltration.
- Sanitization: Robust instructions require the agent to verify all product claims against provided documentation and explicitly forbid the use of unverified certifications, reviews, or metrics from reference materials.
Audit Metadata