browse

Drives a persistent headless Chromium browser for QA testing and site verification.

Updated Jun 22, 2026
One-click install
npx skills add https://github.com/aicodepro/ai-agent-nexi --skill browse-aicodepro
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: browse
Source: https://github.com/aicodepro/ai-agent-nexi/tree/main/agent/skills/gstack/browse
Command: npx skills add https://github.com/aicodepro/ai-agent-nexi --skill browse-aicodepro

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires playwright, bun.

What problem does it solve? Manually verifying that a web feature works—clicking through flows, checking console errors, capturing screenshots for bug reports—is slow and hard to reproduce. This Skill gives an AI agent a persistent headless Chromium browser it can drive with fast (~100ms) commands to navigate pages, interact with elements, and collect evidence automatically. ## Core Features & Use Cases - Interactive element snapshots: Capture the ARIA tree with -i to list clickable elements, including non-ARIA dropdown items via cursor-interactive scanning, then click or fill them by reference. - State verification and diffing: Run snapshot -D to diff page state before and after an action, assert visibility/enabled/checked states, and inspect console and network logs. - Visual evidence and responsive testing: Take annotated screenshots, render mobile/tablet/desktop viewports, test file uploads and dialogs, and diff staging against production. - Use Case: After deploying a login flow, navigate to the page, fill credentials, submit, diff the snapshot to confirm the dashboard appeared, and save an annotated screenshot as proof. ## Quick Start Ask the agent to open your staging site in the browser, test the login flow, and take a 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 with goto, run snapshot -i to list interactive elements, then fill and click elements by their @e references. Use snapshot -D afterward to diff the page state and confirm the expected change occurred.

How do I capture a screenshot for a bug report?

Run snapshot -i -a -o with an output path for an annotated screenshot labeling interactive elements, or screenshot for a plain capture. Pair it with the console command output to attach the JavaScript error log as evidence.

Why does snapshot miss dropdown or autocomplete items?

Dynamically rendered popovers and portals often lack ARIA roles, so the accessibility tree omits them. Use the -C cursor-interactive scan (auto-enabled with -i) to detect clickable divs and popover children that the ARIA snapshot misses.

Can I test responsive layouts across device sizes?

Yes. The responsive command captures mobile, tablet, and desktop screenshots in one call, or set an exact size with viewport (e.g., 375x812) and an optional --scale factor up to 3 for retina output.

Does the browser keep login sessions between commands?

Yes. The Chromium instance is persistent, so cookies, tabs, and login sessions survive between commands. The first call auto-starts the browser in about 3 seconds, then subsequent commands run in roughly 100ms.