mcp

An MCP server for extracting personal data from many services: Amazon order history, Garmin activity stats, Zillow favorites, and more.

Quickstart

Run Chrome Fleet, note its service address (e.g. http://192.168.1.2:8300), set it as CHROMEFLEET_URL, then start this MCP server:

export CHROMEFLEET_URL=http://192.168.1.1:8300
docker run -p 23456:23456 -e CHROMEFLEET_URL ghcr.io/remotebrowser/mcp-getgather

Alternative ways to run it:

services:
  remotebrowser-mcp:
    image: ghcr.io/remotebrowser/mcp-getgather
    ports:
      - "23456:23456"
    environment:
      - CHROMEFLEET_URL
    restart: unless-stopped

export CHROMEFLEET_URL=http://192.168.1.1:8300
podman run -p 23456:23456 -e CHROMEFLEET_URL ghcr.io/remotebrowser/mcp-getgather

You'll need uv with Python >= 3.11. After cloning this repo:

uv run -m uvicorn getgather.main:app --port 23456

Connect to MCP clients

Standard config works with most tools:

{
  "mcpServers": {
    "remotebrowser-mcp": {
      "url": "http://127.0.0.1:23456/mcp"
    }
  }
}

Use the Claude Code CLI to add the MCP server:

claude mcp add --transport http remotebrowser-mcp http://localhost:23456/mcp

Follow the MCP install guide, use the standard config above.

Follow the MCP install guide, use the standard config above.

Go to Program in the right sidebar -> Install -> Edit mcp.json. Use the standard config above.

Follow the MCP install guide, use the standard config above.