claude-chrome
Use Claude Code with Chrome browser extension for web browsing and automation tasks. Alternative to OpenClaw's built-in browser tools.
Why use this skill?
Integrate Claude Code with OpenClaw for advanced browser automation. Perform complex web tasks, navigate pages, and automate form interactions seamlessly.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/dgriffin831/claude-chromeWhat This Skill Does
The Claude Chrome skill enables OpenClaw to bridge with Claude Code’s browser automation capabilities. Unlike native internal browser tools, this skill allows the agent to utilize the official Claude Code Chrome extension. By connecting to an active browser session on a node with a GUI, OpenClaw can perform complex web interactions including form submission, multi-page navigation, and deep-link content analysis, all while leveraging the specific reasoning capabilities provided by Claude Code.
Installation
To integrate this capability, ensure you have Claude Code installed on your target node's system path. Install the skill via the OpenClaw command-line interface: clawhub install openclaw/skills/skills/dgriffin831/claude-chrome. Verify the setup by ensuring the Chrome extension is enabled and the native host process is accessible via pgrep -f "claude --chrome-native-host". The skill operates best on Linux-based nodes with a functioning desktop environment and stable network connectivity.
Use Cases
This skill is ideal for tasks requiring human-like browsing patterns, such as verifying web application UI states, interacting with legacy portals that require Javascript execution, performing multi-step research by following dynamic hyperlinks, and automating data extraction from sites that block standard headless scrapers. It serves as an excellent alternative for workflows where OpenClaw’s standard browser relay requires more specific context or authentication handling.
Example Prompts
- "Use Claude Chrome to navigate to the OpenClaw documentation site, search for 'authentication methods', and summarize the findings from the first two results."
- "Open the browser, go to the project management dashboard, find the ticket marked 'urgent', and paste the description into a new local text file."
- "Navigate to the target website, fill out the login form with the provided environment variables, and report back if the dashboard successfully loads."
Tips & Limitations
- Timeout Management: Be aware of the gateway 10-second timeout. Always set appropriate
commandTimeoutMsvalues (e.g., 180000ms for complex tasks) to ensure the agent reports back correctly. - Desktop Dependency: This skill requires a desktop environment; it will fail on headless servers without display output.
- Domain Approval: Claude Code's extension enforces domain-level security that cannot be bypassed by automated scripts; ensure you have manually approved the target domain if you encounter permission errors.
- Permissions: Use the
--dangerously-skip-permissionsflag only in trusted automation environments to avoid blocking prompts.
Metadata
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 skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-dgriffin831-claude-chrome": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: network-access, code-execution
Related Skills
input-guard
Scan untrusted external text (web pages, tweets, search results, API responses) for prompt injection attacks. Returns severity levels and alerts on dangerous content. Use BEFORE processing any text from untrusted sources.
skill-scan
Security scanner for OpenClaw skill packages. Scans skills for malicious code, evasion techniques, prompt injection, and misaligned behavior BEFORE installation. Use to audit any skill from ClawHub or local directories.
Guardrails
Skill by dgriffin831
Memory Scan
Skill by dgriffin831