Mac App
The Trove Mac app is a native macOS menu bar application that runs connectors locally on your machine. It pulls content from your reading sources — blogs, newsletters, Hacker News, Readwise, bookmarks, and more — and syncs it into your Trove account.
What It Does
Section titled “What It Does”- Discovers connectors from the trove-connectors marketplace
- Runs sync locally on your Mac using Node.js — your data never passes through a third-party server
- Handles authentication for sources that require login (opens a browser for you to sign in)
- Tracks sync history so you can see what was pulled and when
Requirements
Section titled “Requirements”- macOS 13 (Ventura) or later
- Node.js (v18+) — the app uses Node.js to run connector scripts
Install
Section titled “Install”Download the latest release from GitHub:
1. Sign in
Section titled “1. Sign in”Open the app from your menu bar and sign in with your Trove account. This links the Mac app to your cloud knowledge base.
2. Enable connectors
Section titled “2. Enable connectors”Switch to the Discover tab to browse all available connectors. Enable the sources you want to sync. Connectors are grouped by category:
| Category | Examples |
|---|---|
| Blogs | a16z, Daring Fireball, Simon Willison, First Round Review |
| Reading | Hacker News, Readwise, Substack, RSS feeds, Pinboard |
| Social | LinkedIn feed, Twitter/X bookmarks |
| Media | YouTube subscriptions, podcast feeds |
| Local | Obsidian vault, browser history, Safari reading list |
3. Authenticate sources
Section titled “3. Authenticate sources”Some connectors require you to log in to the source. Click Authenticate on the connector detail screen — the app opens a browser window where you sign in normally. Cookies are saved locally for future syncs.
4. Sync
Section titled “4. Sync”Click Sync Now on any connector to pull in the latest content. You can also configure automatic syncing from the settings.
How It Works
Section titled “How It Works”The Mac app doesn’t scrape websites from a remote server. Instead:
- It downloads connector scripts from the trove-connectors repo
- Runs them as local Node.js processes on your machine
- Connectors fetch content (using your saved auth cookies when needed)
- Documents are pushed to the Trove cloud API where they’re indexed with vector embeddings
- Content becomes searchable via MCP tools, the web app, and the GraphQL API
This local-execution model means connectors can access sources that require browser authentication without storing your credentials on a server.
Connector Development
Section titled “Connector Development”Want to build your own connector? Check out the connectors repo and the building a connector guide.
Next Steps
Section titled “Next Steps”- Connect an AI assistant to search your synced content
- Browse your data in the web app
- Build a custom app using the GraphQL API