browse

Controls a persistent headless Chromium browser for QA testing, screenshots, and page interaction.

Updated Aug 9, 2026
One-click install
npx skills add https://github.com/raghavbadhwar/rstack --skill browse-raghavbadhwar
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: browse
Source: https://github.com/raghavbadhwar/rstack/tree/main/browse
Command: npx skills add https://github.com/raghavbadhwar/rstack --skill browse-raghavbadhwar

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires playwright.

What problem does it solve? Manually verifying that a web page loads, a form submits, or a deployment works requires opening a browser and clicking through flows by hand. This Skill gives an AI agent a persistent headless Chromium session so it can navigate pages, interact with elements, capture screenshots, and assert page state as evidence. ## Core Features & Use Cases - Page interaction and inspection: Navigate URLs, click, fill, select, hover, and read text, HTML, links, forms, console output, and network requests via element refs from ARIA snapshots. - Visual and state evidence: Take plain or annotated screenshots, responsive multi-viewport captures, before/after snapshot diffs, and environment comparisons between staging and production. - Real-world use case: After deploying a login flow, run the Skill to open the page, fill credentials, submit, diff the snapshot to confirm the dashboard appeared, and attach an annotated screenshot plus console log to a bug report. ## Quick Start Ask the agent to open your staging site in the browser, test the login 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 page with a headless browser from an AI agent?

Use the browse daemon to navigate to a URL, run snapshot -i to list interactive elements, then click or fill elements by their refs. State persists between commands, so cookies and login sessions carry across steps.

How do I take annotated screenshots for a bug report?

Run snapshot -i -a -o with an output path to capture a labeled screenshot, or use the screenshot command for a plain capture. Combine with the console command to attach JavaScript error logs as evidence.

Does the headless browser keep login sessions between commands?

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

Why does snapshot -i miss dropdown or autocomplete items?

Dynamically rendered popovers and portals may not appear in Playwright's accessibility tree. Use snapshot -C to scan for cursor-pointer and click-handler elements outside the ARIA tree, then interact via the @c refs it returns.

Can I compare staging and production versions of a site?

Yes. The diff command accepts two URLs and compares them directly. You can also use snapshot -D after an action to see a unified diff of exactly what changed on the page.

What setup is required before running browse commands?

The Skill checks for a compiled binary and, if missing, runs a one-time setup script that builds the daemon with Bun. If Bun is not installed, the setup installs it with a checksum-verified install script.