What problem does it solve? Manually driving a browser for repetitive tasks like form submission, data extraction, login flows, and web app testing is slow and error-prone. This Skill gives an AI agent a structured command-line workflow to navigate pages, interact with elements via compact refs, capture screenshots, and manage authenticated sessions. ## Core Features & Use Cases - Snapshot-and-Interact Workflow: Take interactive snapshots that assign refs like @e1 to elements, then click, fill, select, and scroll using those refs, keeping context usage low. - Authentication & Session Management: Save credentials in an encrypted auth vault, persist cookies and localStorage across runs, and run isolated parallel sessions for concurrent scraping or testing. - Capture, Diffing & Debugging: Take full-page or annotated screenshots, export pages as PDF, record video, diff page states or screenshots, and profile performance with Chrome DevTools traces. - Use Case: Imagine you need to verify a signup flow on a staging site. Open the page, snapshot the form, fill each field by ref, submit, wait for the redirect, and diff the snapshot to confirm the success page rendered. ## Quick Start Use the agent-browser skill to open https://example.com/login, fill in the email and password fields, submit the form, and take a screenshot of the resulting page.
