Give your LLM more DSLs

>having LLMs author prompts for themselves (free-form that is, not DSPy-style) is disastrous

They've been perfectly capable at managing their own subagents since opus 4.8 and really good at it since fable.