CLI Reference

The Gambiarra CLI provides commands for managing hubs, rooms, and participants.

Commands

serve

Start a hub server.

gambiarra serve [options]

Options:

Option	Description	Default
--port, -p	Port to listen on	3000
--mdns	Enable mDNS auto-discovery	false

Example:

gambiarra serve --port 3000 --mdns

create

Create a new room.

gambiarra create [options]

Options:

Option	Description	Default
--hub	Hub URL	Auto-discover or http://localhost:3000

Example:

gambiarra create
# Output: Room created! Code: ABC123

join

Join a room with your LLM endpoint.

gambiarra join <room-code> [options]

Options:

Option	Description	Default
--endpoint, -e	LLM endpoint URL	Required
--model, -m	Model name	Required
--nickname, -n	Your nickname	Random
--hub	Hub URL	Auto-discover

Example:

gambiarra join ABC123 \
  --endpoint http://localhost:11434 \
  --model llama3 \
  --nickname alice

list

List available rooms.

gambiarra list [options]

Options:

Option	Description	Default
--hub	Hub URL	Auto-discover

Example:

gambiarra list
# Output:
# Available rooms:
#   - ABC123 (3 participants)
#   - XYZ789 (1 participant)

Supported Providers

Gambiarra works with any OpenAI-compatible API:

Provider	Default Endpoint
Ollama	http://localhost:11434
LM Studio	http://localhost:1234
LocalAI	http://localhost:8080
vLLM	http://localhost:8000