browser-setup-devtools

Browser automation setup (DevTools MCP first) Principles Keep prompts minimal; do as much as possible with tools and commands. Always attempt Chrome DevTools MCP first; only fall back to the browser extension when DevTools MCP cannot be used. Workflow Ask: "Do you have Chrome installed on this computer?" If no or unsure: Offer to open the download page yourself and do it if possible. Provide a clickable link: https://www.google.com/chrome/ Continue after installation is confirmed. Check DevTools MCP availability: Call chrome-devtools_list_pages . If pages exist, select one with chrome-devtools_select_page . If no pages, create one with chrome-devtools_new_page (use https://example.com ) and then select it. If DevTools MCP calls fail: Ask the user to open Chrome and keep it running. Retry chrome-devtools_list_pages . If it still fails, ensure opencode.jsonc includes mcp.control-chrome with command ['chrome-devtools-mcp'] and ask the user to restart OpenWork/OpenCode. Retry the DevTools MCP check. If DevTools MCP is ready: Offer a first task ("Let's try opening a webpage"). If yes, use chrome-devtools_navigate_page or chrome-devtools_new_page to open the URL and confirm completion. Fallback only if DevTools MCP cannot be used: Check availability with browser_version or browser_status . If missing, run npx @different-ai/opencode-browser install yourself. Open the Extensions page yourself when possible: macOS: open -a "Google Chrome" "chrome://extensions" Windows: start chrome://extensions Linux: xdg-open "chrome://extensions" Tell the user to enable Developer mode, click "Load unpacked", and select ~/.opencode-browser/extension , then pin the extension. Re-check availability with browser_version . Offer a first task and use browser_open_tab . Response rules Keep each user prompt to one short sentence when possible. Use direct offers like "I can open Chrome now" and follow with tool actions. Always present links as clickable URLs.