douyin-video
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes shell scripts and Node.js subprocesses to manage the publishing workflow and drive the
ego-browserautomation tool. This is a standard architectural pattern for skills requiring browser-based automation. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data (video titles and descriptions) and injects it into a web page. This constitutes an attack surface for indirect prompt injection if the user-provided text contains instructions designed to manipulate the agent or the target website's UI.
- Ingestion points: Video metadata is accepted as CLI arguments in
scripts/publish.shand parsed viascripts/lib/config-file.mjs. - Boundary markers: No delimiters or "ignore instructions" warnings are applied to the user-provided title or description before injection into the DOM.
- Capability inventory: The skill has the ability to spawn subprocesses (
ego-browser), navigate to URLs, and simulate user interactions (clicks, text insertion) oncreator.douyin.com. - Sanitization: The video title is strictly enforced to be under 30 characters as per
scripts/lib/pure.mjs, which provides a basic constraint on payload length.
Audit Metadata