AI-powered browsing memory for Chrome.
Auto-saves context, reaps stale tabs, and lets you search your history the way you think.
Website · Chrome Web Store · Report a Bug · Request a Feature
You open tabs with a purpose — then forget why. You hoard 40, 80, 120 tabs because closing one might lose something important. Tab suspenders don't help; they just freeze the clutter.
TabBrain runs quietly in the background. Every page you visit gets tracked with its full context — when you opened it, how you got there, what you were working on. When a tab goes stale, TabBrain closes it for you, because it already saved the context. When you need something back, describe it in plain English and TabBrain finds it.
- Live tab dashboard — All open tabs grouped by domain or window with real-time activity tracking
- AI-powered summaries — Each page gets a short summary and intent label so you can scan sessions at a glance
- Smart tab reaping — Automatically closes tabs you haven't touched, with configurable thresholds and domain allowlists
- Semantic search — Find pages by describing what they were about, not just by title or URL
- Session timeline — History organized by work sessions, not a flat list
- Navigation paths — Visualize how you moved through pages in a session, from start to finish
- Ask your memory — Chat-style interface: "What was I researching yesterday about auth?" gets grouped, summarized answers
- Cloud sync — Sign in to sync your browsing memory across devices
- MCP integration — Connect TabBrain to Claude Code or other AI tools so agents can query your browsing context
This repo is the public issue tracker and community hub for TabBrain. The source code is maintained in a private repository.
Use this space to:
- Report bugs you encounter while using the extension
- Request features you'd like to see in future versions
- Ask questions about setup, configuration, or usage
- Share feedback on your experience
- Install TabBrain from the Chrome Web Store
- Click the TabBrain icon in your toolbar to open the popup
- Browse normally — TabBrain tracks everything in the background
- Use Search or Ask to find anything from your browsing history
TabBrain's AI features (summaries, intent labels, memory chat) work in two ways:
- Cloud mode — Sign in with your email. AI runs through TabBrain's servers (included with Pro)
- Direct mode — Bring your own Claude API key for local enrichment (works on Free tier)
Connect TabBrain to your coding workflow:
claude mcp add tabbrain -- node /path/to/tabbrain/mcp/server.tsThen ask Claude Code things like: "Use TabBrain to show what I was reading about Next.js middleware yesterday."
Does TabBrain read my page content? TabBrain extracts a short text snippet from each page for search purposes. Full page content is not stored. AI summaries are generated from the title, URL, and snippet only.
What data leaves my device? Nothing, unless you sign in for cloud sync. With cloud sync enabled, visit metadata (URLs, titles, timestamps, summaries) syncs to TabBrain's servers. Page content stays local.
Can I exclude specific sites? Yes. Add domains to the reaper's allowlist in settings to prevent auto-closing. TabBrain still tracks visits for those domains but won't close the tabs.
What happens if I uninstall? Your local browsing data (IndexedDB) is deleted by Chrome automatically. If you have cloud sync, your data remains on the server and will restore when you reinstall and sign back in.
TabBrain is proprietary software. See LICENSE for details.