What problem does it solve? Browser automation typically requires heavyweight runtimes like Puppeteer or MCP servers that spawn separate browser instances and consume large context. This Skill connects directly to your existing Chrome session via the DevTools Protocol, letting you navigate, click, fill forms, and extract data with simple one-shot CLI commands. ## Core Features & Use Cases - Direct Chrome Control: Navigate pages, click elements, fill inputs, press keys, and take screenshots against your running Chrome with its existing credentials and extensions. - Deterministic Multi-Page Targeting: Friendly target names like [target:green-dog] pin commands to specific tabs, avoiding index drift in multi-page workflows. - Structured Data Extraction: Evaluate JavaScript, capture accessibility tree snapshots, and get machine-readable output with --json for programmatic parsing. - Use Case: Automate a login flow by navigating to a page, filling username and password fields, clicking submit, waiting for the dashboard text to appear, and saving a confirmation screenshot. ## Quick Start Ask the agent to install chrome-devtools-cli, enable Chrome remote debugging, then navigate to a URL and take a screenshot of the page.
