A high-performance browser automation tool for web testing, data extraction, and interactive form filling.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agent-browser-2
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install agent-browser-2 using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Browser Automation (agent-browser)?
agent-browser is a versatile tool designed to facilitate seamless web interactions through an AI-friendly interface. It enables developers to perform complex tasks such as navigating websites, taking snapshots of interactive elements, and executing precise clicks or form fills. This skill is a core component of Openclaw Skills, providing the bridge between programmatic commands and live web environments.
By utilizing a unique reference system for elements, it simplifies the process of DOM interaction, making it more resilient and easier to script than traditional selector-heavy methods. Whether you are building scrapers, automated testers, or workflow agents, this tool offers the reliability and depth needed for professional-grade browser control.
Browser Automation (agent-browser) Use Cases
- Automated web application testing and QA validation.
- High-scale data extraction and web scraping from dynamic sites.
- Generating visual documentation via screenshots, PDFs, and video recordings.
- Automating repetitive form submissions and multi-step authentication flows.
- Testing geolocation and responsive design across different viewports.
How Browser Automation (agent-browser) Works
- Initialize the session by opening a specific URL using the open command.
- Generate a snapshot of the page to identify interactive elements and receive unique reference IDs.
- Execute interactions like clicking, filling, or hovering by targeting the specific reference IDs provided in the snapshot.
- Monitor page state changes or wait for specific network idle conditions to ensure synchronization.
- Capture results through data extraction, screenshots, or state saves for future sessions.
Browser Automation (agent-browser) Setup
Install the necessary dependencies and ensure your environment is configured for Openclaw Skills. Set up the executable path if using a custom browser version.
# Set the executable path (optional)
export AGENT_BROWSER_EXECUTABLE_PATH="/path/to/chrome"
# Quick test to open a site
agent-browser open https://example.com
For persistent sessions, you can define a default session name:
export AGENT_BROWSER_SESSION="mysession"
Browser Automation (agent-browser) Data Schema & Taxonomy
The tool organizes data primarily through snapshots and state files. Interaction relies on a transient reference system (@e1, @e2) that maps to the current DOM state.
| Data Type | Format | Description |
|---|---|---|
| Snapshots | JSON/Text | Lists interactive elements with semantic roles and reference IDs. |
| Auth State | JSON | Saves cookies and localStorage to bypass login in subsequent runs. |
| Media | PNG/PDF/WebM | Visual outputs from screenshots, document exports, or video recordings. |
| Network | JSON | Logs of intercepted requests and mocked responses. |
Browser Automation (agent-browser) Advanced Features
- Multi-session support for parallel browser instances and isolated contexts within Openclaw Skills.
- Full network interception and request mocking capabilities for robust testing.
- Integrated video recording and trace generation for deep debugging.
- Semantic locators (find by role, text, or label) as an alternative to reference IDs.
- Device emulation including viewport sizing, geolocation, and color scheme settings.
SKILL.md
Loading
Related Openclaw Skills
Lobster Farmer Feeder
A specialized skill for the Lobster Farmer game that automates lobster feeding through a CLI interface using AI model token data.
SEO Suite
A comprehensive SEO agent that bridges the gap between technical diagnostics and strategic content planning for organic growth.
SaaS Decomposer
A powerful engine for decomposing existing SaaS services into AI agent workflows to establish a clear internalization and cost-saving roadmap.
Review Manager
Automate review collection, sentiment analysis, and AI-generated responses across multiple platforms.
Weather Query
A professional-grade weather integration for AI agents providing real-time conditions and up to 15-day forecasts via the QWeather API.
Querit Advanced Search Skill
A high-performance web search skill for AI agents using the Querit API with precision filtering capabilities.
Featured*
