What problem does it solve? Manually verifying that a web feature works—clicking through flows, checking console errors, capturing screenshots—is slow and hard to automate from an AI coding session. This Skill provides a fast persistent headless browser (~100ms per command) so an agent can navigate pages, interact with elements, and gather visual evidence of bugs. ## Core Features & Use Cases - Page Interaction & State Assertions: Navigate URLs, click, fill forms, upload files, handle dialogs, and assert element states (visible, enabled, checked) via accessibility-tree snapshots with @e refs. - Visual Evidence & Diffing: Take annotated or responsive screenshots, save PDFs, diff snapshots before/after actions, and compare staging vs production pages. - Session Persistence & Handoff: Cookies, tabs, and login state persist across commands; import cookies from real Chromium browsers, and hand off to a visible Chrome window for CAPTCHAs or MFA. - Use Case: After deploying a login flow, ask the agent to open the site, fill credentials, submit, verify the dashboard appears, and produce an annotated screenshot as bug-report evidence. ## Quick Start Use the browse skill to open https://staging.myapp.com, log in with test credentials, and take a screenshot of the dashboard.
