chromectx

For those who routinely end up with 90+ Chrome tabs. Search all your open tabs by title or full page content, save entire windows to JSON, restore them later, and expose everything as RAG context for AI tools via MCP.

# Fast — searches tab titles and URLs only
chromectx search 'figma design system'

Instantly finds the right tab across 90+ open tabs. No page fetching — results appear in milliseconds.

# Deep — fetches and indexes full page content
chromectx search 'max connections pool timeout' --deep

You read something 20 tabs ago about connection pool limits causing timeouts in production — but which tab was it? The title just says "Configuration" or "Advanced Settings." --deep fetches every open page's content and searches the actual text, so you find it without retracing your steps.

And even better: make your open browser tabs RAG context for Claude, Cursor, and other AI tools via the Model Context Protocol.

npx chromectx install

This one command:

1. Checks your environment (macOS, Chrome, automation permissions)
2. Installs chromectx globally
3. Offers to configure the MCP server for your AI tools (Claude Desktop, Cursor, VS Code, etc.)

If you prefer to set things up yourself:

npm i -g chromectx          # install the CLI globally
chromectx setup             # configure MCP for your AI tools

Or add the MCP server config directly:

{
  "mcpServers": {
    "chromectx": {
      "command": "npx",
      "args": ["-y", "chromectx@latest", "mcp"]
    }
  }
}

• macOS (uses AppleScript/JXA to communicate with Chrome)
• Google Chrome
• Node.js >= 18
• Automation permission: your terminal must be allowed to control Chrome

(System Settings > Privacy & Security > Automation)

| Command | Description | | -----------