cloning-video-ads

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external content such as video files, images, and URLs for reference ad analysis.
  • Ingestion points: Workflow Step 1 and Step 2 in SKILL.md describe reading reference ads and product images from user-supplied files or URLs using video_analysis and image_analysis tools.
  • Boundary markers: The workflow includes mandatory human-in-the-loop checkpoints at Step 7 (blueprint approval) and Step 8 (storyboard approval) where the user must review and approve generated plans before any final content generation occurs.
  • Capability inventory: The skill possesses the ability to generate media using the video_generate tool in Step 10, which uses the synthesized prompts derived from the reference material.
  • Sanitization: Instructions in Step 4 and Step 5 guide the agent to create independent descriptions and blueprints, effectively filtering the raw external input through the agent's own reasoning and formatting before it is used for generation.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 11:23 PM
Security Audit — agent-trust-hub — cloning-video-ads