marketing-paid-ads
Pass
Audited by Gen Agent Trust Hub on Aug 3, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches the
agent-browserskill from Vercel Labs' official GitHub repository, a well-known and trusted source. \n- [COMMAND_EXECUTION]: Uses theagent-browserCLI to automate research on major advertising platforms including Meta, Google, and TikTok. \n- [PROMPT_INJECTION]: The skill ingests untrusted data from competitor landing pages, creating an indirect prompt injection surface (Category 8). \n - Ingestion points:
SKILL.md(Research Mode: Step 4 and 5) \n - Boundary markers: None explicitly defined \n
- Capability inventory: File system writes (screenshots) and network operations via
agent-browser\n - Sanitization: No explicit content filtering \n
- Note: This is a low-risk finding consistent with the skill's primary research purpose.
Audit Metadata