Struct TokenEconomy 

pub struct TokenEconomy {
    pub avg_injected_tokens: Option<f64>,
    pub avg_capsules: Option<f64>,
    pub skip_rate: Option<f64>,
    pub overhead_ratio: Option<f64>,
}

C4 — token economy from context.injected events.

Fields

avg_injected_tokens: Option<f64>

Average used_tokens across context.injected events in the window that carry the field. None when no event carries it (old-style).

avg_capsules: Option<f64>

Average capsule_count across events in the window that carry the field. None when no event carries it.

skip_rate: Option<f64>

Skip-rate (skipped / served); None until C7 adds context.served.

overhead_ratio: Option<f64>

F3 — overhead ratio: brain-injected tokens ÷ total run tokens, averaged over runs in the window that carry both context.injected used_tokens AND run.finished total_prompt_tokens (or equivalent).

Currently None because the pipeline does not yet record total_prompt_tokens in run.finished events. When that field is added, this metric will be computed from it. The fixture-level guarantee (brain overhead ratio falls on larger tasks) is proven in the f3_overhead_ratio_falls_on_larger_task unit test in kimetsu_agent::pipeline using adaptive_budget + simulated totals.

Trait Implementations

impl Clone for TokenEconomy

fn clone(&self) -> TokenEconomy

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for TokenEconomy

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

Formats the value using the given formatter.

impl Serialize for TokenEconomy

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

where __S: Serializer,

Serialize this value into the given Serde serializer.