marketing-video
Pass
Audited by Gen Agent Trust Hub on Aug 3, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill identifies a dependency on the agent-browser tool from Vercel Labs for conducting competitive research and analysis.
- [COMMAND_EXECUTION]: Instructions specify the use of standard npm and npx commands to install and configure required browser automation tools.
- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by ingesting content from external websites. (1) Ingestion points: External URLs are accessed using a browser tool during research. (2) Boundary markers: No specific delimiters are used to separate external content from instructions. (3) Capability inventory: The agent can write strategy and script files to local directories. (4) Sanitization: Content retrieved from the web is not explicitly sanitized.
Audit Metadata