seek-engine

Pass

Audited by Gen Agent Trust Hub on Nov 9, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill retrieves assets from well-known public services:
  • Fetches font families and font files from Google Fonts (fonts.googleapis.com).
  • References elevation data tiles from AWS S3 buckets (s3.amazonaws.com/elevation-tiles-prod).
  • [COMMAND_EXECUTION]: Local binaries are used for rendering and media processing:
  • Spawns ffmpeg for video concatenation, audio mixing, and encoding.
  • Spawns Chrome/Chromium via Node.js and the Chrome DevTools Protocol (CDP) for frame capture.
  • [DYNAMIC_EXECUTION]: The skill uses browser automation to execute code during the render process:
  • Utilizes page.evaluate and Runtime.evaluate to seek through animation timelines and retrieve scene information from the browser DOM.
  • [INDIRECT_PROMPT_INJECTION]: The renderer supports data-driven animations via external JSON files, which is a common pattern for video templates:
  • Ingestion points: render.mjs (reads files specified via --data).
  • Boundary markers: Absent; data is injected into the global window.DATA object.
  • Capability inventory: Scripts can spawn subprocesses (ffmpeg, chrome) and perform file system writes.
  • Sanitization: Relies on standard browser security context.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 9, 2026, 03:32 PM