image-to-code

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill defines a process where the agent generates and analyzes images based on user prompts, then implements code based on that analysis. This creates a surface for indirect prompt injection, as instructions could be embedded in the user's design requirements or in the visual elements the agent is instructed to extract text from.
  • Ingestion points: User prompts and text/content extracted during the deep analysis of generated images (SKILL.md).
  • Boundary markers: No explicit delimiters or instructions to ignore embedded commands within the ingested data were found.
  • Capability inventory: The skill is designed to implement frontend code, involving file-write operations (SKILL.md).
  • Sanitization: There are no documented steps for sanitizing or validating user-provided content before it is processed for code generation.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 01:59 PM
Security Audit — agent-trust-hub — image-to-code