video

Pass

Audited by Gen Agent Trust Hub on Jul 14, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill instructs the agent to read external context files like '.agents/product-marketing.md' before starting tasks. This creates a surface for indirect prompt injection where untrusted data could influence the agent's behavior.
  • Ingestion points: '.agents/product-marketing.md', '.claude/product-marketing.md', and 'product-marketing-context.md' in SKILL.md.
  • Boundary markers: No delimiters or ignore instructions are provided for these inputs.
  • Capability inventory: Capability to execute shell commands for setup and generate code for video frames.
  • Sanitization: No filtering or validation of ingested content is mentioned.
  • [EXTERNAL_DOWNLOADS]: The skill recommends installing 'hyperframes' via npm and using 'create-video' via npx. These are legitimate tools for programmatic video generation.
  • [COMMAND_EXECUTION]: Contains instructions for running project initialization and installation commands, such as 'npm install hyperframes' and 'npx create-video@latest'.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 14, 2026, 07:25 AM
Security Audit — agent-trust-hub — video