Localbot Ctl

Name: Localbot Ctl
Rating: 3.5 (1 reviews)
Author: githabideri

By githabideri 👁 18 views ▲ 0 votes

This is a plugin for OpenClaw to control secondary locally running agent or localbots.

GitHub

Configuration Example

{
  "plugins": {
    "load": {
      "paths": ["/path/to/localbot-ctl"]
    },
    "entries": {
      "localbot-ctl": { "enabled": true }
    }
  }
}

README

# localbot-ctl

OpenClaw plugin for controlling local LLM inference via `/lb*` chat commands.

## Commands

| Command | Description |
|---------|-------------|
| `/lbh` | Help — show available commands |
| `/lbs` | Status — active endpoint, model, session tokens |
| `/lbm` | Models — list available models with specs |
| `/lbm <alias>` | Switch model on llama-cpp server |
| `/lbn` | New session — reset LocalBot context |
| `/lbe` | Endpoints — show all inference backends |
| `/lbp` | Performance — benchmark active endpoint |

## Installation

1. Clone to your workspace plugins directory
2. Add to OpenClaw config:

```json
{
  "plugins": {
    "load": {
      "paths": ["/path/to/localbot-ctl"]
    },
    "entries": {
      "localbot-ctl": { "enabled": true }
    }
  }
}
```

3. Restart gateway

## Configuration

The plugin reads from:
- `config/inference-endpoints.json` — Endpoint definitions
- `config/localbot-models.json` — Model metadata (speeds, context, aliases)

## Requirements

- llama-cpp server (or vLLM/Ollama) running
- LocalBot agents configured in OpenClaw
- Matrix rooms with LocalBot access

## License

MIT

## Changelog

See [CHANGELOG.md](./CHANGELOG.md)

voice