ClawKit Logo
ClawKitReliability Toolkit
Back to Registry
Official Verified browser automation Safety 3/5

agent-browser

Browser automation for AI agents via inference.sh. Navigate web pages, interact with elements using @e refs, take screenshots, record video. Capabilities: web scraping, form filling, clicking, typing, drag-drop, file upload, JavaScript execution. Use for: web automation, data extraction, testing, agent browsing, research. Triggers: browser, web automation, scrape, navigate, click, fill form, screenshot, browse web, playwright, headless browser, web agent, surf internet, record video

Why use this skill?

Automate web browsing, scraping, and interaction with the Agentic Browser skill. Master form filling, navigation, and testing for your AI agents.

skill-install — Terminal

Install via CLI (Recommended)

clawhub install openclaw/skills/skills/xyny89/agentic-browser-0-1-2
Or

What This Skill Does

The agent-browser skill provides a robust bridge between AI agents and the web via inference.sh. Built on top of Playwright, it enables automated browser control, allowing agents to navigate, interact, and extract information from web pages. The skill utilizes a unique @e (element) reference system, which maps interactive DOM components to simple handles, making complex multi-step web interactions programmatic and reliable. Whether you are managing browser sessions, filling out complex web forms, or scraping data, this skill facilitates seamless communication between the agent's logic and the internet.

Installation

To integrate this skill into your environment, use the OpenClaw Hub command: clawhub install openclaw/skills/skills/xyny89/agentic-browser-0-1-2. After installation, ensure you have initialized your inference.sh CLI environment with infsh login. This verifies your authentication and readies your local environment to handle headless browser sessions via the cloud.

Use Cases

  • Automated Data Extraction: Scrape tables, lists, and dynamic content from sites that require JavaScript to render.
  • Complex Form Workflows: Automate repetitive logins, sign-ups, or survey entries across multiple domains.
  • E2E Testing: Build autonomous test suites that verify UI functionality by navigating pages and triggering interaction sequences.
  • Research and Monitoring: Monitor competitor pricing, news updates, or social sentiment by instructing an agent to navigate to specific pages periodically and parse the content.

Example Prompts

  1. "Open https://www.example-store.com, find the search bar, search for 'wireless headphones', and give me the prices of the first 5 results."
  2. "Go to the login page at https://app.service.com, fill out the email with '[email protected]' and the password, then click the 'Submit' button. Let me know when you reach the dashboard."
  3. "Navigate to https://techcrunch.com, take a full-page screenshot of the homepage, and extract the top three headlines into a JSON format."

Tips & Limitations

  • Session Lifecycle: Always remember to use the close function when your task is complete to free up cloud resources and ensure you receive your video recording (if enabled).
  • Snapshotting: The DOM changes frequently in modern web apps. If you notice your @e refs aren't working after a button click, call the snapshot function to re-index the page elements.
  • Limitations: While highly capable, this skill relies on the inference.sh cloud infrastructure; verify your internet connectivity and ensure the targeted sites do not have aggressive bot-detection measures that might block headless browsers.

Metadata

Author@xyny89
Stars879
Views3
Updated2026-02-11
View Author Profile
AI Skill Finder

Not sure this is the right skill?

Describe what you want to build — we'll match you to the best skill from 16,000+ options.

Find the right skill
Add to Configuration

Paste this into your clawhub.json to enable this plugin.

{
  "plugins": {
    "official-xyny89-agentic-browser-0-1-2": {
      "enabled": true,
      "auto_update": true
    }
  }
}

Tags(AI)

#browser-automation#web-scraping#playwright#web-agent
Safety Score: 3/5

Flags: network-access, code-execution, data-collection