Browser Automation Expert
Pass
Audited by Gen Agent Trust Hub on Jun 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions guide the agent to create and run local Python scripts for automation tasks using frameworks like Playwright and Selenium.
- [PROMPT_INJECTION]: The skill interacts with external, untrusted web content, creating a vulnerability to indirect prompt injection where web-based instructions could influence agent behavior.
- Ingestion points: Web content retrieved via Puppeteer, Playwright, or custom automation scripts.
- Boundary markers: None specified to isolate external data from instruction context.
- Capability inventory: Includes browser-side JavaScript execution and local Python script execution.
- Sanitization: No validation or sanitization of external web data is described in the provided instructions.
Audit Metadata