Chrome DevTools MCP for browser automation. Five servers configured for parallel agents. Auto-select available server, cleanup when done.
/plugin marketplace add zeulewan/zeul-claude-plugins/plugin install chrome-devtools@zeul-claude-pluginsThis skill inherits all available tools. When active, it can use any tool Claude has access to.
Five MCP servers configured: chrome-devtools, chrome-devtools-2, chrome-devtools-3, chrome-devtools-4, chrome-devtools-5
list_pages - Initialize browser, see open pagesnew_page(url) - Open URL (auto-selects new page)take_screenshot(filePath) - Capture to fileclose_page(pageIdx) - Close page by indexCheck availability - Server is available if list_pages returns only about:blank:
Check servers in order until one is available:
1. mcp__chrome-devtools__list_pages
2. mcp__chrome-devtools-2__list_pages
3. mcp__chrome-devtools-3__list_pages
4. mcp__chrome-devtools-4__list_pages
5. mcp__chrome-devtools-5__list_pages
→ Only "about:blank"? Use that server's tools
→ Has other pages? Check next server...
→ All busy? Cleanup one (close_page for index > 0), then use it
Why multiple servers? Parallel Task agents sharing ONE server race-condition on "selected page". Each server = own browser = no conflicts.
Chrome-devtools-mcp has NO idle timeout. Manually free servers:
# After completing work:
1. list_pages → note indices
2. close_page(pageIdx=N) for each page except about:blank (index 0)
"Browser already running" / stale locks:
rm -f ~/.cache/chrome-devtools-mcp/chrome-profile/Singleton*
Race condition with parallel agents:
Page not loading:
wait_for(text) after navigationBoth servers use --headless --isolated:
--headless = No visible browser window--isolated = Temp profile, auto-cleaned on closeAdd more servers if needed:
claude mcp add chrome-devtools-3 --scope user -- npx chrome-devtools-mcp@latest --headless --isolated
This skill should be used when the user asks to "create a slash command", "add a command", "write a custom command", "define command arguments", "use command frontmatter", "organize commands", "create command with file references", "interactive command", "use AskUserQuestion in command", or needs guidance on slash command structure, YAML frontmatter fields, dynamic arguments, bash execution in commands, user interaction patterns, or command development best practices for Claude Code.
This skill should be used when the user asks to "create an agent", "add an agent", "write a subagent", "agent frontmatter", "when to use description", "agent examples", "agent tools", "agent colors", "autonomous agent", or needs guidance on agent structure, system prompts, triggering conditions, or agent development best practices for Claude Code plugins.
This skill should be used when the user asks to "create a hook", "add a PreToolUse/PostToolUse/Stop hook", "validate tool use", "implement prompt-based hooks", "use ${CLAUDE_PLUGIN_ROOT}", "set up event-driven automation", "block dangerous commands", or mentions hook events (PreToolUse, PostToolUse, Stop, SubagentStop, SessionStart, SessionEnd, UserPromptSubmit, PreCompact, Notification). Provides comprehensive guidance for creating and implementing Claude Code plugin hooks with focus on advanced prompt-based hooks API.