runtimeInjectsToolDeclarations static method

bool runtimeInjectsToolDeclarations(
  1. ModelType? modelType, {
  2. ModelFileType? fileType,
})

Whether this model's format has its tool declarations injected by the runtime/SDK, so InferenceChat must not weave its own tools prompt. The model-derived default for InferenceChat.runtimeInjectsToolDeclarations; an engine that injects tools natively (e.g. the web Prompt API) overrides it per-chat. Keyed off the format, not a hardcoded model.

Implementation

static bool runtimeInjectsToolDeclarations(
  ModelType? modelType, {
  ModelFileType? fileType,
}) => FunctionCallFormatFactory.create(
  modelType,
  fileType: fileType,
).runtimeInjectsToolDeclarations;