Optimize token usage by routing each goal to the cheapest capable AI agent CLI. - View it on GitHub