Module llm 

LLM provider abstraction.

A provider takes a transcript plus tool specs and returns either free-form content, structured tool calls, or both. The trait is the only thing the agent depends on; everything beyond it (HTTP, retries, mocking) lives in adapters.

Re-exports

pub use anthropic::AnthropicProvider;

pub use openai::OpenAiProvider;

Modules

anthropic

Anthropic Messages API adapter.

mock

Deterministic LLM for tests and offline development.

openai

OpenAI-compatible chat-completions adapter.

search

Tool search engine: keyword-based discovery of deferred tools.

Structs

Completion

One step of model output.

ModelPricing

Per-million-token pricing for one model. USD.

RetryPolicy

Retry policy for transient LLM provider failures (network timeouts, 5xx).

StructuredRequest

Request for a structured JSON response conforming to a JSON schema.

TokenUsage

Token usage data from an LLM response.

ToolCall

A structured request to invoke one of the registered tools.

ToolSpec

JSON-schema description of a tool, sent verbatim to the model.

Traits

LlmProvider

Functions

context_window_tokens_for_model

Returns the context window size in tokens for the given model.

default_compact_threshold_chars

Compute the default compaction character-count threshold for a model.

pricing_for

Returns pricing for a model by looking it up in the bundled providers.toml. Returns None if the model is not listed or has no pricing field.

Type Aliases

StreamSender

Channel sender for streaming partial tokens during a streaming LLM call. Each String is a delta chunk (partial token) emitted by the provider.