diverge

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze user-provided design references, creating a surface where malicious instructions could be embedded in the external data.
  • Ingestion points: The agent is directed to analyze the vibe, palette, typography, and other metadata from external design references.
  • Boundary markers: No specific delimiters or instructions to ignore embedded commands within the reference material are provided in SKILL.md.
  • Capability inventory: The skill generates HTML/CSS snippets and interfaces with other internal design-related skills like frontend-design.
  • Sanitization: No sanitization or validation of the external reference content is specified.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 03:06 PM
Security Audit — agent-trust-hub — diverge