an.semantic.prompt

The an iterate system prompt, generated from the vocabulary registry.

ADR 0003 decision 6: the prompt is one of the three surfaces generated from the registry, never written by hand. vocabulary_prompt() renders the IR’s fields and every registered name — action and entity kinds, motion and expression presets, camera moves, easings, methods by aspect — each with its one-sentence description and its params. The protocol around it (the patch operations, the path syntax, the editing rules) is an iterate’s own and is passed in as preamble and postamble.

>>> text = vocabulary_prompt()
>>> "push_in" in text and "tween" in text and "linear" in text
True

Functions

iterate_prompt(*, preamble, postamble)	preamble + the generated vocabulary + postamble (the an iterate protocol).
vocabulary_prompt(*[, kinds])	The IR fields and every registered name, as prompt text.

an.semantic.prompt.iterate_prompt(*, preamble, postamble)

preamble + the generated vocabulary + postamble (the an iterate protocol).

Return type:

str

an.semantic.prompt.vocabulary_prompt(*, kinds=(('action', "Action kinds (an action's `kind`)"), ('entity', "Entity kinds (an entity's `kind`)"), ('motion_preset', 'Motion presets (`play` by name; `args` are the parameters shown)'), ('expression_preset', "Expression presets (an `expression`'s `preset`, a dialogue line's `emotion`)"), ('camera_move', 'Camera moves (`camera: {move: …}`)'), ('easing', "Easings (a tween's `easing`)")))

The IR fields and every registered name, as prompt text.

Return type:

str