browse

Controls a persistent headless Chromium browser for QA testing, screenshots, and element state assertions.

Updated Apr 5, 2026
One-click install
npx skills add https://github.com/Simon-YHKim/eject-button --skill browse-simon-yhkim
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: browse
Source: https://github.com/Simon-YHKim/eject-button/tree/main/.claude/skills/browse
Command: npx skills add https://github.com/Simon-YHKim/eject-button --skill browse-simon-yhkim

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires bun, and includes references (resource) components.

What problem does it solve? Manually verifying that a web page loads correctly, a user flow works, or a bug is reproducible requires opening a browser, clicking through steps, and capturing evidence by hand. This Skill automates that loop with a persistent headless Chromium instance, producing annotated screenshots, console logs, and element state assertions as bug evidence. ## Core Features & Use Cases - Interactive page control: Navigate any URL, click, fill forms, upload files, handle dialogs, and switch tabs using @ref selectors from accessibility tree snapshots. - Visual evidence capture: Take annotated screenshots with labeled element overlays, responsive layout screenshots (mobile/tablet/desktop), and before/after snapshot diffs. - State assertions and inspection: Assert visibility, enabled, checked, and focused states; inspect console errors, network requests, cookies, and computed CSS. - Use Case: After deploying a login flow, run a scripted sequence that fills credentials, submits the form, diffs the page state, and saves an annotated screenshot proving the dashboard renders. ## Quick Start Ask the AI to open your staging site in the browser, walk through the signup flow, and take an annotated screenshot of the result.

Frequently Asked Questions about browse

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I test a web user flow with a headless browser?

Navigate to the page with goto, run snapshot -i to list interactive elements with @e refs, then fill inputs and click buttons by ref. Use snapshot -D afterward to diff what changed and is visible to assert the success state.

How to take annotated screenshots for bug reports?

Run snapshot -i -a -o /tmp/annotated.png to save a screenshot with red overlay boxes and ref labels on each interactive element. Combine with console and network output to capture full bug evidence.

Can a headless browser handle CAPTCHA or MFA login?

Headless automation cannot solve CAPTCHAs or multi-factor prompts. Use the handoff command to open a visible Chrome window at the current page, let the user complete the step, then resume with all cookies and state preserved.

Does headless browser testing support responsive layouts?

Yes. The responsive command captures screenshots at mobile (375x812), tablet (768x1024), and desktop (1280x720) sizes in one call. You can also set a custom viewport with viewport WxH before taking a screenshot.

Why do element refs stop working after navigation?

Refs are invalidated on navigation because the accessibility tree is rebuilt for the new page. Run snapshot again after every goto or page change to get fresh @e refs before interacting.