Dashboard UI

TribalMind includes a built-in web dashboard for managing your knowledge base visually. It’s useful when you want to browse, search, or clean up memories without using the CLI.

Launching the dashboard

tribal ui

This starts a local web server and opens your browser to http://localhost:7484.

Options

# Custom port
tribal ui --port 8080

# Don't auto-open browser
tribal ui --no-browser

Press Ctrl+C to stop the server.

Features

Assistants view

The main page lists all Backboard assistants in your account. Each assistant represents a project (or the global assistant). You can:

• See the assistant name and ID
• Click into an assistant to view its memories
• Delete assistants you no longer need

Memory management

Within an assistant, you can:

• Browse all stored memories in a scrollable list
• Search for specific memories
• Create new memory entries directly from the UI
• Delete individual memories

Activity log

The dashboard includes an activity log that shows a chronological feed of all memory interactions — remembers, recalls, and forgets. You can:

• Browse recent activity across all assistants
• Filter by assistant to focus on a specific project
• See which assistant each action targeted

This gives you a bird’s-eye view of how your team and agents are using the knowledge base.

When to use the dashboard vs CLI

Use case	Recommended
Quick recall while coding	CLI (tribal recall)
Storing knowledge from terminal	CLI (tribal remember)
Browsing and auditing all memories	Dashboard
Bulk cleanup of outdated memories	Dashboard
Checking what agents have stored	Dashboard
Reviewing activity across assistants	Dashboard (Activity Log)
Debugging assistant configuration	Dashboard

Architecture

The dashboard is a React + TypeScript frontend served by a FastAPI backend, all bundled with the TribalMind Python package. When you run tribal ui:

1. A FastAPI server starts on the configured port
2. The server serves the pre-built React frontend as static files
3. API requests from the frontend are proxied to the Backboard API using your stored credentials
4. Your browser opens to localhost:7484 (or your custom port)

No data leaves your machine except the Backboard API calls — the same calls the CLI makes.