browser-ui-qa
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill identifies the processing of untrusted external content (web pages and test fixtures) as a potential attack surface. It provides explicit defensive instructions for the agent to treat repository content as data rather than instructions and specifically to flag any 'ignore previous instructions' patterns encountered during testing.
- [EXTERNAL_DOWNLOADS]: The skill references the axe-core accessibility engine from the well-known Cloudflare CDN (cdnjs.cloudflare.com) and the playwright-core package for UI auditing. These resources are standard in the industry and originate from trusted service providers.
- [COMMAND_EXECUTION]: The workflow involves shell commands for the browse CLI (Browserbase) and node. It also manages local file operations, such as creating directories for screenshots and report artifacts in the .context/ folder, which is typical for browser automation tools.
- [DYNAMIC_EXECUTION]: JavaScript is dynamically executed in the browser context via browse eval to perform automated audits, such as extracting accessibility violations or measuring performance metrics. This behavior is necessary for the intended use case and is performed using established automation frameworks.
Audit Metadata