frontend-image-generation

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [Indirect Prompt Injection] (SAFE): The skill interpolates user-controlled data into image generation prompts and API URLs, creating a surface for potential instruction injection in downstream processing.
  • Ingestion points: Variable interpolation in SKILL.md for avatar seeds (seed=${name}) and DALL-E prompts (prompt: '...[concept]...').
  • Boundary markers: No delimiters or explicit instructions are provided to the agent to ignore control characters in user inputs.
  • Capability inventory: The skill has high privileges including Bash (*) and file system access (Read, Write, Edit).
  • Sanitization: No evidence of input validation or prompt escaping is present in the skill snippets.
  • [Credential Exposure] (SAFE): The code snippets reference the OPENAI_API_KEY environment variable rather than hardcoding secrets.
  • [Command Execution] (SAFE): While the skill requests broad Bash (*) permissions, the documentation and code blocks do not contain instructions to execute arbitrary or dangerous commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:46 PM