Enum ToolChoice 

pub enum ToolChoice {
    Auto,
    None,
    Required,
    Tool(String),
}

How the model should route tool selection for the current turn. Mapped to each provider’s wire field by chat_request_body:

variant	OpenAI	Anthropic
Auto	"auto" (or omitted)	{"type":"auto"} (or omitted)
None	"none" — model MUST NOT call a tool	(degrades: tools field dropped)
Required	"required" — model MUST call at least one tool	{"type":"any"}
Tool(name)	{"type":"function","function":{"name":name}} — forced	{"type":"tool","name":name}

The None variant has no direct Anthropic equivalent — Anthropic forces tool consideration whenever tools is non-empty. We approximate it by dropping the tools field entirely for that turn; the model can’t call what it doesn’t know about.

Variants

Auto

None

Required

Tool(String)

Implementations

impl ToolChoice

pub fn parse(s: &str) -> Self

Parse from the bootstrap.yaml string form. Accepts "auto", "none", "required", "tool:<name>". Empty / unknown strings degrade to Auto for forward compat — callers that want strict validation can match before calling.

Trait Implementations

impl Clone for ToolChoice

fn clone(&self) -> ToolChoice

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ToolChoice

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

Formats the value using the given formatter.

impl Default for ToolChoice

fn default() -> ToolChoice

Returns the “default value” for a type.

impl PartialEq for ToolChoice

fn eq(&self, other: &ToolChoice) -> 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 ToolChoice