Configuration

moldable onboard

moldable gateway status --config /path/to/config.json5

​

Where the config lives

• macOS/Linux default: ~/.moldable/gateway/config.json5
• Windows default: %APPDATA%\\Moldable Gateway\\config.json5

​

What you typically configure

• Gateway auth: a local token used by the CLI and any enabled HTTP endpoints.
• Channels: enable Telegram/WhatsApp and provide credentials.
• AI backend: choose which backend the gateway will call for responses.
• Pairing/allowlists: control who is allowed to message the gateway.
• Optional HTTP endpoints: enable OpenAI-compatible HTTP endpoints if you need them.

​

Minimal example

{
  "gateway": {
    "auth": { "mode": "token", "token": "replace-me" }
  },
  "channels": {
    "telegram": {
      "enabled": true,
      "bot_token": "<telegram-bot-token>",
      "require_mention": true
    }
  },
  "ai": {
    "default_adapter": "ai-server",
    "adapters": [
      { "type": "ai-server", "name": "ai-server", "base_url": "http://127.0.0.1:39200" }
    ]
  }
}