ad-creative

Pass

Audited by Gen Agent Trust Hub on Aug 19, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest untrusted data from customer reviews and ad comments to ground creative generation, which creates a potential surface for indirect injection.
  • Ingestion points: The skill reads customer-generated text from inputs/reviews/ and inputs/comments/ as specified in SKILL.md.
  • Boundary markers: The instructions do not define specific delimiters or instructions for the agent to differentiate instructions from the ingested review content.
  • Capability inventory: The agent has access to advertising platform tools and local shell execution for media generation.
  • Sanitization: No specific text sanitization or validation routines for external inputs are described.
  • [EXTERNAL_DOWNLOADS]: The documentation references several well-known generative AI services and provide examples for installation and API usage.
  • References official APIs and documentation for Google Gemini, OpenAI, ElevenLabs, and ByteDance (Seedance).
  • Provides instructions for installing the Voicebox tool via GitHub and the Remotion framework via NPM (generative-tools.md), both of which are established development projects.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 19, 2026, 02:49 PM