Automates web interactions and page analysis using deterministic Playwright scripts within the Claude Code CLI.
The Browser Control skill empowers Claude Code with full web automation capabilities by leveraging bundled Playwright scripts. It allows the AI to perform complex browser tasks such as navigating to URLs, capturing DOM snapshots, simulating user actions like clicking and typing, and taking screenshots for visual verification. Designed with performance in mind, it includes token optimization strategies like jq filtering for DOM snapshots and action batching, while maintaining security through a granular domain-based access control policy for read and write operations.
Key Features
01Deterministic DOM snapshotting and element analysis
020 GitHub stars
03Simulated user interactions including click, type, and key presses
04Visual page captures and screenshot reporting for UI verification
05Full browser automation using Playwright and Chromium
06Granular security policies for domain-level read/write access control
Use Cases
01Extracting structured data and DOM information from web pages for research
02Visual regression testing and UI state verification using automated screenshots
03Automated end-to-end testing of web applications directly from the terminal
