Browser Automation Tool

OpenClacky has built-in browser automation that lets the Agent control your local Chrome or Edge browser — clicking, filling forms, taking screenshots, and scraping page content — including pages that require login or JavaScript rendering.


How It Works

The Agent connects to your browser's remote debugging port via the Chrome DevTools Protocol (CDP), using your existing login sessions and cookies. No re-login needed.

Connection is auto-detected at runtime, in priority order:

  1. DevToolsActivePort file — if the browser wrote a DevToolsActivePort file on startup, OpenClacky reads the port and WebSocket path directly for zero-config connection
  2. TCP port scan — scans ports 9222–9224 to detect a listening browser
  3. Auto-connect fallback — if neither works, falls back to --autoConnect to launch the browser

Supported Browsers

Browser macOS Linux Windows (WSL)
Google Chrome
Microsoft Edge
Chrome Beta / Canary -

Before You Start: Launch Browser in Debug Mode

The Agent needs the browser started with remote debugging enabled. Choose the command for your system:

macOS

Chrome:
bash
/Applications/Google\ Chrome.app/Contents/MacOS/Google\ Chrome \
--remote-debugging-port=9222 --no-first-run

Edge:
bash
/Applications/Microsoft\ Edge.app/Contents/MacOS/Microsoft\ Edge \
--remote-debugging-port=9222 --no-first-run

Linux

Chrome:
bash
google-chrome --remote-debugging-port=9222 --no-first-run

Edge:
bash
microsoft-edge --remote-debugging-port=9222 --no-first-run

Windows (WSL)

WSL cannot directly launch Windows-side browsers. Run the following in a Windows PowerShell window instead:

Chrome:
powershell
& "C:\Program Files\Google\Chrome\Application\chrome.exe" `
--remote-debugging-port=9222 --no-first-run

Edge:
powershell
& "C:\Program Files (x86)\Microsoft\Edge\Application\msedge.exe" `
--remote-debugging-port=9222 --no-first-run

💡 Tip: Save this as a .ps1 script so you can launch it quickly each time.


Quick Setup: browser-setup Skill

Use the built-in browser-setup skill to guide you through the setup and verify the connection:

/browser-setup

Or just say:

setup browser
help me configure the browser

The skill auto-detects your environment (macOS / Linux / WSL), verifies the connection, and saves the config to ~/.clacky/browser.yml.

See Built-in Skills → browser-setup.


Config File

Settings are stored in ~/.clacky/browser.yml, usually generated automatically by the browser-setup skill.

# ~/.clacky/browser.yml
user_data_dir: /path/to/UserDataDir   # optional: manually specify user data directory

Troubleshooting

Agent can't find the browser / connection failed

Check:
1. The browser was launched with --remote-debugging-port=9222 (a normally opened browser won't work)
2. Port 9222 is not blocked by a firewall or used by another process
3. Run browser-setup doctor for automated diagnostics

Does Edge work in WSL?

Yes. In WSL, OpenClacky automatically queries powershell.exe to locate the Windows-side Edge/Chrome user data directory, then reads the DevToolsActivePort file to connect — as long as you launched the browser on the Windows side with debug mode enabled.

Browser version requirements

Chrome / Edge v146 or later is required. The /json/version HTTP endpoint was removed in v146; only WebSocket connections are supported now. OpenClacky handles this correctly.