Struct TokenUsage 

pub struct TokenUsage {
    pub input_tokens: u64,
    pub output_tokens: u64,
    pub total_tokens: u64,
}

Aggregated token usage for one worker attempt, normalized across worker kinds (agent-block agent.run return / subprocess declared mapping / operator self-report). Field names follow the Anthropic wire convention (input_tokens / output_tokens) that agent-block already normalizes OpenAI-style responses into.

Every wire field is optional on the way in (TokenUsageWire): a reporter that only knows one axis (a harness completion notice that surfaces a single token total, an API response that omits the total) still lands a usable usage record instead of being dropped. The stored shape stays the closed 3-field triple — missing splits read as 0, and a missing total_tokens is derived as input + output on decode.

Fields

input_tokens: u64

Prompt-side tokens consumed, summed across the attempt’s turns. 0 when the reporter did not split its total.

output_tokens: u64

Completion-side tokens produced, summed across the attempt’s turns. 0 when the reporter did not split its total.

total_tokens: u64

input + output (kept explicit because some producers report a total that includes cache-read/creation tokens the two split fields don’t cover). Derived from the splits when the reporter omits it.

Implementations

impl TokenUsage

pub fn from_parts( input_tokens: Option<u64>, output_tokens: Option<u64>, total_tokens: Option<u64>, ) -> Option<Self>

Build a usage record from three independently-optional parts, applying the same normalization the wire decode does: absent splits read as 0, and an absent total is derived as input + output.

None only when the reporter carried no token axis at all — the caller then records no usage rather than a zeroed one.

Shared by the hand-rolled extractors that read usage out of a worker’s raw payload (agent-block’s agent.run return, subprocess usage_ptr declarations) so every axis applies one normalization rule.

Trait Implementations

impl Clone for TokenUsage

fn clone(&self) -> TokenUsage

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for TokenUsage

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for TokenUsage

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Eq for TokenUsage

impl From<TokenUsageWire> for TokenUsage

fn from(w: TokenUsageWire) -> Self

Converts to this type from the input type.

impl JsonSchema for TokenUsage

fn schema_name() -> Cow<'static, str>

The name of the generated JSON Schema.

fn schema_id() -> Cow<'static, str>

Returns a string that uniquely identifies the schema produced by this type.

fn json_schema(generator: &mut SchemaGenerator) -> Schema

Generates a JSON Schema for this type.

fn inline_schema() -> bool

Whether JSON Schemas generated for this type should be included directly in parent schemas, rather than being re-used where possible using the $ref keyword.

impl PartialEq for TokenUsage

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

Equality operator ==.

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

Inequality operator !=.

impl Serialize for TokenUsage

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl StructuralPartialEq for TokenUsage