Struct ModelTurnInput 

pub struct ModelTurnInput {
    pub system_prompt: Option<String>,
    pub messages: Vec<ChatMessage>,
    pub tools: Vec<ToolSpec>,
    pub tool_choice: ToolChoice,
    pub parallel_tool_calls: Option<bool>,
}

Fields

system_prompt: Option<String>

Optional system prompt (already composed: spec_snapshot system_prompt

• driver.append_system_prompt). None ⇒ no system message sent.

messages: Vec<ChatMessage>

Full conversation history for this turn. AgentLoopHarness appends each Assistant / Tool message as the loop progresses so the model retains its own prior reasoning across tool round-trips.

tools: Vec<ToolSpec>

Tool specs available this turn. Sourced from ToolRuntime::specs() so adding / removing a tool changes one place. Empty Vec ⇒ no tools advertised (final-answer-only mode).

tool_choice: ToolChoice

How the model should pick (or skip) tools. Defaults to Auto. Set via AgentLoopHarness::with_tool_choice from bootstrap.driver.native_model.tool_choice.

parallel_tool_calls: Option<bool>

Whether the model may emit multiple tool_use blocks in one response (OpenAI’s parallel_tool_calls). None ⇒ provider default (true for OpenAI). Anthropic is always implicitly multi-tool-capable so this field is OpenAI-only.

Trait Implementations

impl Clone for ModelTurnInput

fn clone(&self) -> ModelTurnInput

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for ModelTurnInput

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl PartialEq for ModelTurnInput

fn eq(&self, other: &ModelTurnInput) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl StructuralPartialEq for ModelTurnInput