image-generation

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data, including text prompts, product URLs, and uploaded images, to generate task handoffs for downstream skills like image-batch-runner.
  • Ingestion points: User-supplied prompts, product URLs, and external image assets identified in SKILL.md.
  • Boundary markers: The instructions lack explicit delimiters or instructions for the agent to ignore potentially malicious embedded commands within the ingested content.
  • Capability inventory: The skill identifies input types and task classes to generate artifacts for other runners and interacts with the postplus CLI.
  • Sanitization: No validation or filtering mechanisms are described for the external URLs or user-provided prompt text.
  • [COMMAND_EXECUTION]: The skill documentation explicitly instructs the agent to execute CLI commands and scripts as part of its workflow.
  • It details the usage of postplus quote confirm --json --challenge-file <challenge.json> to resolve quote-confirmation challenges.
  • The instructions mandate choosing and running the smallest matching command or workflow based on the provided user input.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 03:07 PM
Security Audit — agent-trust-hub — image-generation