ad-render

Warn

Audited by Gen Agent Trust Hub on Jul 17, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes multiple shell commands to manage its workflow, including kai-config for configuration and npx remotion for the video rendering process.
  • [EXTERNAL_DOWNLOADS]: In Step 4, the skill performs an npm install within a dynamically generated project directory, fetching external Node.js packages from the npm registry.
  • [REMOTE_CODE_EXECUTION]: The skill runs npx remotion render to execute the video generation logic within the scaffolded project, which involves running code and dependencies fetched at runtime.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting untrusted data from creative brief files in ~/.kai-marketing/briefs/ and the --copy argument. This data is interpolated into the agent's context and scene breakdowns without the use of boundary markers or sanitization, potentially allowing malicious ad copy to influence agent behavior.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 17, 2026, 02:28 PM