Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
- Added skill.json file to define or enhance skill metadata. - No changes made to functionality or documentation in SKILL.md.
What This Skill Does
Controls Chrome via the Chrome DevTools Protocol through an MCP server. Supports navigating pages, interacting with DOM elements, running JavaScript, capturing screenshots, and analyzing network requests. The browser starts automatically and maintains a persistent profile across tool calls.
Gives agents direct access to a live browser without requiring a separate Playwright or Selenium setup.
When to use it
- Debug a broken form that fails silently in production
- Scrape data from a JavaScript-rendered page
- Trace slow network requests to find a bottleneck
- Automate a multi-step login flow for testing
- Inspect DOM structure to fix a layout regression
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Why is the checkout button not responding on my e-commerce page?
- 1Call navigate_page to open the checkout page
- 2Call wait_for to confirm the page is fully loaded
- 3Call take_snapshot to map the page structure and locate the button uid
- 4Call evaluate_script to check for JavaScript errors or missing event listeners on the button
- 5Report the console errors and handler state back to the user
Identifies that a missing script dependency is preventing the click handler from attaching
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.