bx-playwright-browsing

Pass

Audited by Gen Agent Trust Hub on Oct 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONPERSISTENCEDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed for web browsing and scraping, creating an ingestion point for untrusted data from external websites.
  • Ingestion points: page.visit(), page.content(), and various finders/locators in SKILL.md.
  • Boundary markers: None specified in the provided instructions.
  • Capability inventory: The skill can perform file writes (fileWrite), browser-side script execution (evaluate), and chainable actions like click and fill.
  • Sanitization: No explicit sanitization of scraped content is described.
  • [PERSISTENCE]: The skill includes instructions for creating persistent scheduled tasks using the BoxLang scheduler (scheduler.task().everyHour()) and refers to OS-level cron for automation.
  • [DYNAMIC_EXECUTION]: Provides the ability to execute Javascript strings within the browser context via page.evaluate( js, arg ). It also facilitates the execution of BoxLang scripts (.bxs files).
  • [COMMAND_EXECUTION]: Instructions describe the use of command-line tools for environment setup (bxPlaywright install) and script execution (boxlang jobs/capture.bxs).
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 14, 2026, 08:51 PM