Chrome DevTools Skill
Control a real Chrome browser with Claude. The Chrome extension runs inside an authenticated browser session, allowing interaction with sites the user is already logged into -- Gmail, Google Docs, Notion, CRMs, and more.
Overview
Two integration paths exist: Claude Code + Chrome Extension (terminal-based browser control via claude --chrome) and Chrome DevTools MCP (26 browser automation tools via Model Context Protocol). Both operate on a real Chrome instance with existing logins, requiring no re-authentication. The Chrome extension also supports workflow shortcuts, scheduled recurring tasks, and multi-tab coordination independently of Claude Code.
Quick Reference
| Integration | Setup | Best For |
|---|
| Claude Code + Extension | claude --chrome | Terminal-based browser control |
| Chrome DevTools MCP | claude mcp add chrome-devtools --scope user npx chrome-devtools-mcp@latest | Programmatic automation via MCP tools |
| Capability | Tools/Methods |
|---|
| Navigation | Navigate URLs, create/close/switch tabs, resize windows |
| Input | Click, type, fill forms, drag, hover, press keys, upload files |
| Extraction | Read page content, DOM state, console logs, network requests |
| Debugging | Screenshots, DOM snapshots, evaluate JavaScript, performance tracing |
| Emulation | Device emulation, viewport resizing |
| Recording | Record browser interactions as GIFs, save workflow shortcuts |
| Scheduling | Workflow shortcuts with daily/weekly/monthly/annual triggers |
| Requirement | Minimum |
|---|
| Google Chrome | Latest stable |
| Claude in Chrome Extension | 1.0.36+ |
| Claude Code CLI | 2.0.73+ |
| Claude Plan | Pro, Team, or Enterprise |
| Node.js (DevTools MCP) | 20.19+ |
Not supported: Other Chromium browsers (Brave, Arc, Edge), WSL, headless mode (extension path only; DevTools MCP supports headless), mobile devices.
Common Mistakes
| Mistake | Correct Pattern |
|---|
Running browser automation without --chrome flag | Start Claude Code with claude --chrome or enable Chrome by default via /chrome settings |
| Trying to take over existing tabs | Claude opens new tabs for tasks; it does not hijack tabs you already have open |
| Not handling CAPTCHAs and modal dialogs manually | Claude pauses at CAPTCHAs, login prompts, and modals; dismiss them yourself then tell Claude to continue |
| Using ambiguous instructions for form filling | Be specific about field names, values, and order; ambiguous prompts produce inconsistent results |
| Requesting all console logs without filtering | Specify log patterns or error types; requesting everything floods context with noise |
| Enabling Chrome by default without need | Increases context usage since browser tools are always loaded; use --chrome flag only when needed |
Using DevTools MCP without --scope user | Add --scope user to make the MCP server available across all projects |
Delegation
- Automate repetitive browser workflows across sites: Use
Task agent with Chrome extension to fill forms, extract data, or navigate multi-step flows
- Debug frontend issues with console and network inspection: Use
Explore agent with DevTools MCP to capture console errors, network failures, and DOM state
- Plan complex multi-tab data extraction pipelines: Use
Plan agent to design the workflow steps before executing browser automation
- Record and schedule recurring browser tasks: Use Chrome extension shortcuts with scheduling for daily/weekly automation
References