ai-image-generation

Pass

Audited by Gen Agent Trust Hub on Jul 13, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill documents the installation of the @runcomfy/cli package from the npm registry for interacting with the image generation service.
  • [COMMAND_EXECUTION]: Uses the runcomfy CLI tool to perform text-to-image and image-to-image operations. The skill correctly limits execution scope using the allowed-tools configuration.
  • [PROMPT_INJECTION]: Identifies potential indirect prompt injection surfaces where external data (image URLs and web search) is processed by the AI models. • Ingestion points: External image URLs in image_urls / images fields and web grounding via enable_web_search in SKILL.md. • Boundary markers: The documentation includes a "Security & Privacy" section with explicit instructions for agents to treat external reference assets as untrusted content. • Capability inventory: The agent's capabilities are restricted to executing the runcomfy command. • Sanitization: All user-provided prompts are encapsulated within a JSON object and passed to the CLI via the --input flag, effectively neutralizing shell injection risks.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 13, 2026, 04:10 AM
Security Audit — agent-trust-hub — ai-image-generation