better-browser-use

Automates browser navigation, interaction, and debugging via agent-browser with synced local login state.

Updated Sep 14, 2026
One-click install
npx skills add https://github.com/Suge8/skills --skill better-browser-use-suge8
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: better-browser-use
Source: https://github.com/Suge8/skills/tree/main/operations/better-browser-use
Command: npx skills add https://github.com/Suge8/skills --skill better-browser-use-suge8

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires agent-browser.

What problem does it solve? AI agents often need to operate and debug web pages that require the user's existing login sessions or anti-detection measures, and switching between isolated automation browsers and the user's daily browser breaks sessions and pollutes state. ## Core Features & Use Cases - Dual-lane browser automation: Run agent-browser commands against an isolated Chrome for Testing instance per repository, or switch to a CloakBrowser anti-detection lane with a synced login profile via --login. - Login state synchronization: bbu sync-profile snapshots cookies from the user's Helium browser online (via SQLite backup), transcrypts them to the automation profile, and copies Preferences, Local Storage, and Network data. - Debugging workflow: Structured loop of open, interactive snapshot, console error and network request inspection, actions, and waits, with CDP fallback for shadow DOM, canvas, and cross-origin iframes. - Use Case: While developing a web app, ask the agent to open your dev server, fill a form, click submit, then inspect console errors and failed network requests — all in a browser that already has your site logins. ## Quick Start Ask the agent to open a URL with bbu, take an interactive snapshot, and report any console errors or failed network requests.

Frequently Asked Questions about better-browser-use

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

FAQPage Schema
How do I automate browser actions with login sessions from my daily browser?▼

Run bbu sync-profile to snapshot cookies from your Helium browser via the SQLite backup API and transcrypt them into the dedicated automation profile. Then use bbu --login followed by any agent-browser command to operate with that login state.

How to debug console errors and network requests in an automated browser?▼

Use the default bbu lane, which runs Chrome for Testing with full console, errors, and network access. Run errors --json for reliable output and network requests --status 400-599 to filter failures, adding --filter with a URL substring when multiple tabs are open.

Why are console and errors always empty in the anti-detection browser lane?▼

The --login lane uses CloakBrowser, whose engine by design suppresses console and exception events, so console and errors commands always return empty there. Switch back to the default Chrome for Testing lane for debugging output.

Can browser automation handle shadow DOM, canvas, or cross-origin iframes?▼

Standard element references fail on shadow DOM, canvas, and cross-origin iframes. Use screenshot --annotate with get box to calibrate mouse coordinates, or retrieve the CDP websocket endpoint via get cdp-url and send raw Chrome DevTools Protocol commands.

What are the safety limits of agent-driven browser automation?▼

The eval command is read-only, and the workflow forbids printing cookies or tokens, dumping entire DOM or storage, and executing payments, deletions, message sending, password changes, or production data submissions. Desktop applications are out of scope.