NYC

design-video

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • Indirect Prompt Injection (LOW): The skill is designed to process untrusted video data, which could contain malicious instructions designed to influence the agent's behavior.
  • Ingestion points: The $ARGUMENTS variable is used to pass video content into the multimodal analysis workflow.
  • Boundary markers: Absent. While the input is enclosed in <video> tags, there are no instructions for the agent to ignore potential prompt injections embedded within the video's visual or audio tracks.
  • Capability inventory: The skill has the capability to create directories and write multiple file types (Markdown, HTML, CSS, JavaScript) to the local file system.
  • Sanitization: No sanitization or validation of the instructions derived from the video content is performed before they are used to generate code or documentation.
  • Dynamic Execution (LOW): The skill's primary purpose is to generate and implement functional code (HTML/CSS/JS). While this is the intended behavior, the resulting code is dynamically created from unverified external input, which poses a minor risk if the analysis phase is compromised.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:43 PM