This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
MCP server for accessing browser data (tabs, history, bookmarks) from Chromium-based browsers.
Supports Brave, Chrome, and Chromium.
# Using uvx (recommended)
uvx chromium-sync-mcp
# Or install with pip
pip install chromium-sync-mcpRequires the LevelDB library:
# Ubuntu/Debian
sudo apt-get install libleveldb-dev
# macOS
brew install leveldb
# Fedora
sudo dnf install leveldb-develRelated MCP server: Chrome History & Bookmarks MCP Server
Add to your Claude Code MCP settings:
{
"mcpServers": {
"chromium-sync": {
"command": "uvx",
"args": ["chromium-sync-mcp"]
}
}
}Tool | Description |
| Get open tabs from all synced devices |
| Get open tabs from the local browser session |
| Search browsing history with optional filters |
| Get bookmarks, optionally filtered by folder |
| Search bookmarks by title or URL |
| Select which browser to use (when multiple installed) |
| Manually set the browser profile path |
| Check what data is accessible (for debugging) |
Returns a JSON array of history entries. Supports substring search, regex patterns, and date filtering.
Parameters:
Parameter | Type | Description |
| string | Substring match against URL and title (case-insensitive). Cannot be used with |
| string | Regex match against URL and title. Cannot be used with |
| integer | Maximum results to return. Default: 100 |
| integer | Only return entries from the last N days. |
| string | ISO date or datetime. Only entries on or after this time. |
| string | ISO date or datetime. Only entries before this time. |
Date formats: YYYY-MM-DD or YYYY-MM-DDTHH:MM:SS
Example response:
[
{
"url": "https://github.com/anthropics/claude-code",
"title": "GitHub - anthropics/claude-code",
"visit_time": "2026-01-11T14:30:00",
"visit_count": 5
}
]The server automatically detects installed Chromium-based browsers. If multiple browsers are found, you'll be prompted to select one.
Override auto-detection by setting CHROMIUM_PROFILE_PATH:
export CHROMIUM_PROFILE_PATH=~/.config/google-chrome/DefaultWhen prompted to select a browser, use select_browser with save_default: true to save your preference to ~/.config/chromium-sync/profile.
Browser | Linux | macOS | Windows |
Brave | ✓ | ✓ | ✓ |
Chrome | ✓ | ✓ | ✓ |
Chromium | ✓ | ✓ | ✓ |
This server reads directly from your browser's local profile files:
History: SQLite database
Bookmarks: JSON file
Synced Tabs: LevelDB (contains tabs from all your synced devices)
No authentication or network requests required.
If you're running on a headless server and need to enter your Chrome sync passphrase, use the chromium-sync-setup command. It launches a browser in a virtual display and provides a secure web URL for remote access.
This is a one-time setup per machine. Once you've entered your passphrase and sync is established, you won't need to run this again.
# If you installed via uvx (recommended)
uvx --with chromium-sync-mcp[setup] --from chromium-sync-mcp chromium-sync-setup
# If you installed via pip
pip install chromium-sync-mcp[setup]
chromium-sync-setupWhat it does:
Starts a virtual X display (Xvnc or Xvfb)
Launches your browser to the sync settings page
Provides a secure HTTPS URL via Cloudflare tunnel
System requirements (one of):
TigerVNC: sudo apt install tigervnc-standalone-server
Or Xvfb + x11vnc: sudo apt install xvfb x11vnc
The script auto-downloads cloudflared and noVNC, so those don't need manual installation.
Apache 2.0
This server cannot be deployed
Web search, data extraction and controlled browser automation for AI assistants via OAuth.
Read and search folders on your own computer with explicit consent. Files are never uploaded.
Mac & Windows: let ChatGPT, Claude & Cursor use your email, calendar, iMessage, Teams, files.
Connect AI agents to your own Chrome with scoped access, approvals, and verified task results.
