Crate llm_api

Expand description

Canonical model invocation and continuation contract. No Agent or transport implementation.

Re-exports§

Completion: Provider-neutral completion returned to Runtime.
CompletionRequest: One logical LLM invocation.
Continuation: Provider-issued replay data attached to exactly one completed model message. Adapters own the format and compatibility key. Callers must not synthesize or edit payloads.
GenerationParameters: Deployment-owned generation settings, never supplied through Agent constraints. None leaves the setting unspecified. Adapters must apply every explicit value or reject it; this basic validation does not establish model/provider support.
Message: Provider-neutral conversation message.
ModelCapabilities: Confirmed model capabilities. Missing information never establishes support.
ModelConstraints: Caller-declared requirements. True requires confirmed support; false imposes no requirement. Adapters must not infer or override these flags from message content.
ModelMode: Logical model mode selected by the product, independent of provider and model identifiers.
ModelProfile: Provider-neutral limits for one frozen logical model route.
TokenUsage: Normalized token accounting reported by an implementation.
ToolCall: Tool call returned by an LLM.
ToolDefinition: A callable function exposed to the model. Execution policy belongs to the caller.
UseCase: Logical model use case resolved by the deployment’s LLM implementation.