Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
- Major restructure: the skill bundle now directly includes the OpenClaw plugin, server, and client SDK for streamlined installation. - Added full server and client implementation (server/, clients/) to the published bundle. - Installation docs updated for ClawHub and workspace environments; clarified plugin registration path requirements. - Node.js 16+ is now required (was 14+). - Removal of legacy plugin entry files; only the new openclaw-plugin/ directory is needed. - Troubleshooting and deployment instructions improved, with explicit steps for npm dependencies.
What This Skill Does
Browser extension and WebSocket server that lets AI agents automate a connected browser. Supports listing and managing tabs, extracting HTML, executing JavaScript in page context, and reading cookies. Works with Chrome, Edge, and Firefox.
Operates within the user's existing browser session, so agents can access authenticated pages without re-implementing login flows.
When to use it
- Scraping content from pages that require login
- Running JavaScript snippets in a live tab to extract dynamic data
- Reading session cookies from authenticated browser tabs
- Opening and navigating URLs as part of an automated research task
- Closing stale tabs based on URL patterns
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Summarize the content of my open GitHub PR tab
- 1Calls js_eyes_get_tabs to list all open browser tabs
- 2Identifies the GitHub PR tab by its URL
- 3Calls js_eyes_get_html with the tab ID to retrieve full HTML
- 4Parses and summarizes the PR description and comments
A summary of the PR title, description, and recent comments
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.