slideshow

Pass

Audited by Gen Agent Trust Hub on Aug 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill reads and processes external content which could contain malicious instructions designed to influence the agent's behavior during the ad generation process.
  • Ingestion points: The skill reads copy-deck.md and brand-profile.json as primary inputs (SKILL.md).
  • Boundary markers: There are no specified delimiters or instructions to ignore embedded commands within the ingested files.
  • Capability inventory: The skill triggers shell command execution via npx remotion for both image and video rendering (SKILL.md).
  • Sanitization: No sanitization or validation of the text content within the copy deck is mentioned before it is passed to the rendering engine.
  • [COMMAND_EXECUTION]: The skill executes shell commands via npx remotion to render assets. It dynamically constructs these commands using file paths and properties derived from the input files, which is a required capability for its stated purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 18, 2026, 07:53 PM
Security Audit — agent-trust-hub — slideshow