agent-browser
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWEXTERNAL_DOWNLOADS
Full Analysis
- EXTERNAL_DOWNLOADS (LOW): The skill references a repository from a trusted source (vercel-labs/agent-browser). Per [TRUST-SCOPE-RULE], these external references are categorized as LOW risk.
- COMMAND_EXECUTION (INFO): The skill documentation describes how to use CLI commands for browser automation (click, fill, open). This is the intended functionality of the skill and does not contain unauthorized or malicious command injection.
- INDIRECT PROMPT INJECTION (LOW): As a tool designed to process web pages and accessibility trees, there is an inherent surface for indirect prompt injection from external websites. However, the skill acts as documentation/bridge and does not implement unsafe logic or lack boundary markers in its current form.
Audit Metadata