guided-testimonial-video
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill fetches content from external, user-provided URLs to extract brand intelligence in Step 3. This ingested content is then used to generate creative specifications and implementation code, which could allow malicious instructions on the remote site to influence the agent's behavior.
- Ingestion points: External brand/marketing URLs provided by the user.
- Boundary markers: None specified; instructions do not include delimiters or warnings to ignore embedded directives in the fetched content.
- Capability inventory: The skill performs file writes (composition.html) and shell command execution (npm install, npx helios).
- Sanitization: No validation or sanitization of the fetched site content is described.
- [DYNAMIC_EXECUTION]: In Step 8, the skill generates a
composition.htmlfile containing HTML, CSS, and JavaScript (Web Animations API). This generated code is subsequently executed and rendered by the Helios CLI in Step 9 to produce the final video file. - [EXTERNAL_DOWNLOADS]: The skill installs and executes third-party Node.js packages (@helios-project/core, @helios-project/renderer, @helios-project/cli) from the NPM registry during its setup and execution phases.
Audit Metadata