agent-browser

Automate browser navigation, form filling, and UI testing with persisted sessions.

20|4|Updated Oct 4, 2025
One-click install
npx skills add https://github.com/coalesce-labs/catalyst --skill agent-browser-coalesce-labs
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/coalesce-labs/catalyst/tree/main/plugins/dev/skills/agent-browser
Command: npx skills add https://github.com/coalesce-labs/catalyst --skill agent-browser-coalesce-labs

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill enables AI agents to perform browser-based tasks like opening pages, filling forms, taking screenshots, and testing UI with persisted sessions.

Core Features & Use Cases

  • Session-driven control: manage named browser sessions that survive across commands for continuity.
  • Navigation & interaction: open URLs, click elements, fill forms, type text, and simulate user actions.
  • Observability & automation: snapshot the page state, capture screenshots, extract text or metadata, and wait for elements or conditions.
  • Use Case: automate end-to-end UI testing, QA verification, data collection from websites, and guided product demos.

Quick Start

Open a browser session named 'task1', navigate to https://example.com, and take a screenshot.

Frequently Asked Questions about agent-browser

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

FAQPage Schema
How do I automate web testing with persistent browser sessions for AI agents?

Persistent browser sessions for AI agents are automated by creating a named CLI session and using agent-browser commands like open, click, and fill. This maintains state across commands for reliable end-to-end UI testing and form filling.

Can AI agents perform headless browser automation and take screenshots during navigation?

Yes, AI agents can perform headless browser automation by using agent-browser commands to navigate URLs, simulate clicks, fill forms, and capture screenshots. The session-driven approach ensures reliable scripted navigation and observability.

What is the best way to set up a browser session for automated UI testing and data collection?

The best way to set up browser sessions for automated UI testing is to initialize a named CLI session, then apply consistent prompts with agent-browser commands. This enables reliable web automation, QA verification, and data extraction.

Does browser automation for AI agents support waiting for elements and extracting page metadata?

Yes, browser automation for AI agents supports waiting for specific page elements and conditions. It enables you to snapshot the page state, extract text or metadata, and capture screenshots for automated QA checks and data collection.

Why do I need consistent prompts when using the agent-browser commands for web automation?

Consistent prompts are required for web automation because the AI agent needs clear, repeatable instructions to execute agent-browser commands accurately for open, click, fill, and screenshot actions across the persistent named session.

Are there limitations when using session-driven browser automation for guided product demos?

Session-driven browser automation for guided product demos requires a named CLI session and consistent prompts. It is limited to scripted navigation and UI testing actions, relying heavily on stable element selectors for reliable execution.