playwright-skill

Pass

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads the 'playwright' package and required browser binaries (Chromium) from official NPM and Playwright repositories during the setup phase.
  • [COMMAND_EXECUTION]: Features a universal executor ('run.js') that uses 'require' to dynamically execute Node.js automation scripts generated by the agent or provided via command-line arguments.
  • [COMMAND_EXECUTION]: Automatically scans common local development ports (e.g., 3000, 3001, 5173, 8080) using Node's 'http' module to identify active targets for browser automation.
  • [COMMAND_EXECUTION]: Supports injecting custom HTTP headers into browser sessions through environment variables ('PW_HEADER_NAME', 'PW_EXTRA_HEADERS'), enabling the identification of automated traffic to backends.
  • [COMMAND_EXECUTION]: Metadata discrepancies exist between the skill's manifest (author: 'ncdevshiv', version: '1.0') and its package configuration (author: 'lackeyjb', version: '4.1.0').
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 21, 2026, 11:40 PM
Security Audit — agent-trust-hub — playwright-skill