browse

What problem does it solve? Manually verifying that a web page loads correctly, that a user flow works end to end, or that a deployment matches staging requires opening a browser, clicking around, and capturing evidence. This Skill gives an AI agent a fast persistent headless Chromium instance (~100ms per command after startup) to navigate, interact, inspect, and screenshot any web page on demand. ## Core Features & Use Cases - Page verification and inspection: Navigate to any URL, extract text/HTML/links/forms, check console errors, network requests, performance timings, and assert element states (visible, enabled, checked, focused). - Interactive flow testing: Snapshot pages to get @e/@c element refs, then click, fill, select, upload files, handle dialogs, and diff before/after snapshots to verify exactly what changed. - Visual evidence and responsive checks: Capture annotated screenshots, PDFs, and mobile/tablet/desktop responsive screenshots, or diff text content between two environments (e.g., staging vs production). - Session reuse: Import cookies from JSON files or from installed Chromium browsers (Chrome, Arc, Brave, Edge, Comet) via a local cookie-picker UI to test authenticated flows. - Use Case: After deploying a new login page, run goto, snapshot -i, fill credentials, click submit, then snapshot -D and is visible ".dashboard" to confirm the flow works, and save an annotated screenshot as bug-report evidence. ## Quick Start Ask the agent to open your app's URL with the browse skill, fill in the login form, submit it, and verify the dashboard appears with a screenshot.