using-chrome-devtoolslisted

Automates headless Chrome via MCP for web scraping, screenshots, testing, and browser interactions. Handles multi-instance server selection to avoid lock conflicts between parallel agents. Triggers when browsing a website, taking a screenshot, scraping a page, testing a web app, automating browser interactions, or when chrome-devtools MCP tools are needed.
Saturate/agents · ★ 0 · AI & Automation · score 66

Install: claude install-skill Saturate/agents

You are helping the user interact with websites using Chrome via the chrome-devtools MCP servers. There are 4 isolated server instances available. You must pick a free one before doing any browser work. ## Progress Checklist Copy this checklist to track your progress through the workflow: ``` Chrome DevTools Progress: - [ ] Step 1: Selected an available server instance - [ ] Step 2: Loaded required MCP tools via ToolSearch - [ ] Step 3: Performed the requested browser task - [ ] Step 4: Cleaned up (closed pages) ``` ## Step 1: Select an Available Server There are 4 MCP server instances — 3 headless and 1 visible (opens a real browser window): | Instance | Tool prefix | Mode | |----------|-------------|------| | 1 | `mcp__chrome-devtools__*` | Headless | | 2 | `mcp__chrome-devtools-2__*` | Headless | | 3 | `mcp__chrome-devtools-3__*` | Headless | | visible | `mcp__chrome-devtools-visible__*` | Visible | **When to use visible:** Use `chrome-devtools-visible` when the user asks to see the browser, needs to debug visually, or wants to watch automation in real-time. Default to headless instances for everything else. **To find a free instance**, check each server's `list_pages` tool (e.g. `mcp__chrome-devtools__list_pages`). Load them via ToolSearch: ``` ToolSearch: "+chrome-devtools list_pages" ``` Then call `list_pages` on each instance. A server is **free** if it only has the default `about:blank` page (or returns no pages). A server is **busy** if it has pages with r