Commit
be75d54
Only run my orchestrator high, and have it delegate to discrete agents across Claude and GPT based on what effort level it thinks is required for that agent’s task.
Sometimes that means Sonnet, sometimes Sol, sometimes Opus, sometimes Terra
I mean, I know what’s configured and available.
I don’t look too closely at what’s actually being used. I don’t think it matters too much - the orchestrator keeps everything pretty consistent anyway.