Skip to main content

​
Prerequisites

  • Python 3.11 or later
  • Node.js 20+ (for the dashboard UI)
  • An API key for at least one LLM provider, or Ollama installed locally

​
1. Clone and install

For the dashboard:

​
2. Create your config

Writes a starter ~/.neuralcleave/config.toml. Edit it to set your provider and key under [models]:
Set your key in the environment:
See Configuration for the full reference, and LLM Providers for every supported model ID prefix.

​
3. Start the gateway

Or, to also open the dashboard in one step:

​
4. Open the dashboard

Or manually run the frontend dev server:
Navigate to http://localhost:3000. The chat page is live.

​
5. Send your first message

From the terminal, without any channel or dashboard:
Or check gateway status first:

​
Next steps

Connect a channel

Route messages from Telegram, Discord, or Slack.

Enable voice

Add wake word detection and local STT/TTS.

Add a skill

Teach the agent a new capability.

Switch providers

Use Ollama for a fully local setup.