Features Overview

Everything Uplink can do — powered by OpenClaw, controlled by you.

Built on OpenClaw

Uplink is designed as a frontend for OpenClaw — the open-source AI agent platform. Every feature in Uplink leverages OpenClaw's capabilities under the hood:

Local-First Architecture

Uplink runs as a Node.js server on localhost:3456. Your conversations, settings, and data live on your machine. The AI connection goes through the OpenClaw gateway running locally, so your prompts and responses never touch third-party servers unless you explicitly configure an external endpoint.

When your internet goes down, Uplink keeps serving. The web UI, settings, and conversation history are all local. Only the AI inference requires a working gateway connection.

Text Chat

The core experience. A clean, responsive chat interface in your browser. Features include:

Satellites

Satellites are independent AI conversations with isolated context. Each satellite maintains its own message history, system prompt, and context window — completely separate from your other conversations. Think of them as parallel chat sessions that don't interfere with each other.

ℹ Satellites ≠ Providers Satellites are session-level separation, not model-level. They all use the same OpenClaw gateway and whatever model it's configured with. They're part of OpenClaw's embed/widget architecture.

Activity Panel

The activity panel gives you a real-time window into what OpenClaw is doing under the hood. Toggle it with the wrench icon (🔧) in the header to see live developer and debug information as your conversation flows.

💡 When to use it The activity panel is especially useful when debugging tool failures, monitoring long-running tasks, or just satisfying your curiosity about how OpenClaw processes your requests.

Voice Chat

Talk to your AI instead of typing. Uplink supports multiple TTS and STT engines — from free cloud services to fully local GPU-accelerated models. See the Voice Chat page for full details.

EngineTypeRuns WhereNotes
ElevenLabsText-to-SpeechCloudHighest quality, API key required
OpenAI TTSText-to-SpeechCloud (paid)High quality, requires API key
Edge TTSText-to-SpeechCloud (free)Microsoft's free TTS API, many voices
PiperText-to-SpeechLocal (CPU)Fast, lightweight, many voices
Coqui XTTSText-to-SpeechLocal (GPU)Best local quality, needs NVIDIA GPU
OpenAI WhisperSpeech-to-TextCloudUses OpenAI API key
GroqSpeech-to-TextCloudFree tier available, fast
faster-whisperSpeech-to-TextLocalFast, accurate, runs on CPU or GPU

End-to-End Encryption

Uplink supports optional AES-256 encryption for all conversations. Enable it on first run or later in Settings → Privacy. When enabled:

ℹ How It Works Your password is used to derive an AES-256 key via PBKDF2 in the browser. Messages are encrypted client-side before being sent to the server. The server only ever sees ciphertext.

Themes

Five built-in themes to match your vibe. Switch instantly in Settings → General. See the Themes page for previews.

ThemeAccent ColorStyle
MidnightCyan (#00f0ff)Dark — the default
DaylightWashed Red (#c84b4b)Light — warm bg (#f5f0eb)
EmberAmber (#f48c06)Deep fire, dark navy bg (#03071e)
ForestSage (#a3b18a)Earthy dark, dark green bg (#131813)
PhantomGray (#a0a0a0)Monochrome grayscale

PWA Support

Uplink is a Progressive Web App. On mobile, add it to your home screen for a native-app experience — full-screen, no browser chrome, offline-capable UI. Works on both iOS Safari and Android Chrome.

Multi-Device Access

Since Uplink is a web server, any device on your network can access it by navigating to the server's IP and port. Conversations sync in real-time via WebSocket. Use it on your desktop, laptop, phone, and tablet simultaneously.

For access outside your home network, see Remote Access.

OpenClaw Gateway

Uplink connects to AI exclusively through the OpenClaw gateway. OpenClaw provides an OpenAI-compatible API that handles model routing, session management, and streaming — Uplink is built to take full advantage of these features.

💡 Why OpenClaw? OpenClaw isn't just an API proxy — it's an AI agent platform with session memory, multi-model routing (satellites), tool execution, and more. Uplink is built to surface all of this through a clean chat interface. Features like the satellite switcher, context tracker, and persistent conversations all depend on OpenClaw's session architecture.
⚠ Other endpoints While the API format is OpenAI-compatible, pointing Uplink at a raw OpenAI endpoint, llama.cpp, or LM Studio will give you basic chat only. Features like satellites, session persistence, context tracking, and gateway status won't work without OpenClaw.

Agents

The Agents Panel lets you manage AI agent sessions, routing, and configuration directly from the Uplink interface. Access it to control how OpenClaw routes requests and configure agent-specific behavior.

Watchdog Auto-Restart

Uplink includes a watchdog process monitor that automatically restarts the server if it crashes or becomes unresponsive. Ensures high availability for long-running deployments.

File Uploads

Attach files directly to your messages for the AI to analyze. Uplink supports images, documents, and spreadsheets.

Push Notifications

Stay informed even when Uplink isn't in the foreground. Web push notifications alert you when new messages arrive — useful when running long tasks or waiting for agent responses.

Missed Messages

Never lose track of messages received while you were away. The missed messages system queues messages that arrived while you were offline and lets you catch up when you return.

Conversation Sharing

Generate public share links for your conversations. Useful for sharing interesting AI interactions, debugging sessions, or collaborative work with others.

Keyboard Shortcuts

Power users can navigate Uplink efficiently with keyboard shortcuts for common actions.