Quickstart

Get a working AI agent on your machine in 5 minutes.

Prerequisites

• Ollama running locally with at least one model pulled:

  ollama pull qwen2.5:7b

• Go 1.22+ (for building from source) or a pre-built vibe binary

Install vibe

From source:

git clone https://github.com/contenox/runtime
cd runtime
go install ./cmd/vibe

Initialize a workspace

mkdir my-agent && cd my-agent
vibe init

This creates .contenox/ with a default config and chain:

.contenox/
├── config.yaml          ← model, backend settings
└── default-chain.json   ← the default task chain

Start chatting

vibe "what is the capital of France?"
# → Paris.

vibe "list files in the current directory" --enable-local-exec
# → the model calls local_shell with `ls` and returns the result

vibe without a subcommand is interactive — type your message and press Enter. Ctrl+D exits.

Run a chain explicitly

vibe exec --chain .contenox/default-chain.json --input-type chat "explain recursion briefly"

Add a remote API as a tool

# US National Weather Service — free, no API key
vibe hook add nws --url https://api.weather.gov --timeout 15000
vibe hook show nws       # lists 60 discovered tools

vibe exec --chain .contenox/chain-nws.json --input-type chat \
  "how many active weather alerts are there right now?"

Next steps

• Core Concepts — understand chains, tasks, and hooks
• Chains reference — write your own chains
• vibe CLI reference — all flags and subcommands