TurboStream Terminal UI

January 29, 2026 · View on GitHub

Keyboard-driven terminal interface for real-time data streaming with AI

A beautiful terminal UI built with Bubble Tea for monitoring high-velocity data streams and analyzing them with LLMs in real-time.

Quick Start Video

📹 Coming Soon: Watch our TUI demo and setup guide on YouTube

Get Started in 3 Steps

1. Start the Backend

The TUI connects to the TurboStream backend. Start it first:

# Clone and run the backend
git clone https://github.com/turboline-ai/turbostream.git
cd turbostream
cp .env.local.example .env.local
# Edit .env.local with your MongoDB and LLM credentials
go run ./cmd/server

Backend runs at http://localhost:7210 by default.

2. Clone & Run the TUI

git clone https://github.com/turboline-ai/turbostream-tui.git
cd turbostream-tui
go run .

Enter your credentials (from backend user registration) and start streaming! 🚀

Features

Real-Time Streaming

WebSocket connection with automatic reconnection
Token-by-token LLM response streaming
Live feed data monitoring
Sub-second latency

Observability Dashboard

Press d to toggle the dashboard:

📹 Coming Soon: Watch the dashboard tour video

Panels:

Stream Health - Connection status, throughput, reconnections
LLM Context - Memory usage, context size, eviction stats
Payload Stats - Size distribution, averages, histograms
LLM Performance - Token counts, TTFT, generation time

Visualization:

Sparkline charts for real-time metric trends
60-second rolling windows
Color-coded indicators (green = good, red = issues)

For detailed metric definitions: DASHBOARD_METRICS_REVIEW.md

Feed Management

Browse available feeds
Subscribe/unsubscribe in real-time
Monitor multiple feeds simultaneously
Custom AI prompts per feed

Key Bindings

Key	Action
`Enter`	Submit / Execute action
`d`	Toggle dashboard view
`q`	Quit application
`↑/↓`	Navigate feed list
`c`	Reconnect WebSocket
`Tab`	Cycle through inputs
`Esc`	Go back / Cancel

📹 Coming Soon: Watch the keyboard shortcuts tutorial

Configuration

Environment Variables

Variable	Description	Default
`TURBOSTREAM_BACKEND_URL`	Backend REST API URL	`http://localhost:7210`
`TURBOSTREAM_WEBSOCKET_URL`	WebSocket endpoint	`ws://localhost:7210/ws`
`TURBOSTREAM_TOKEN`	Pre-configured JWT token	None
`TURBOSTREAM_EMAIL`	Pre-fill login email	None

Example:

export TURBOSTREAM_BACKEND_URL=https://your-backend.railway.app
export TURBOSTREAM_WEBSOCKET_URL=wss://your-backend.railway.app/ws
go run .

Screenshots

Main Feed View

Feed View

Observability Dashboard

📹 Coming Soon: Dashboard walkthrough video

Development

Prerequisites

Go 1.24+
Running TurboStream backend

Build from Source

# Clone
git clone https://github.com/turboline-ai/turbostream-tui.git
cd turbostream-tui

# Install dependencies
go mod download

# Build
go build -o turbostream-tui .

# Run
./turbostream-tui

Project Structure

turbostream-tui/
├── main.go              # Main application entry
├── ws.go                # WebSocket handling
├── pkg/
│   └── api/             # Backend API client
└── README.md