Method used_tokens

atomic-core/src/change/attestation.rs:274–276

Used tokens across all models (input + output only). This is the number that should be displayed as the headline "Tokens" metric — it matches what the user sees in OpenCode and represents actual computation, not cached context.

(&self)

    /// metric — it matches what the user sees in OpenCode and represents
    /// actual computation, not cached context.
    pub fn used_tokens(&self) -> u64 {
        self.models.iter().map(|m| m.used_tokens()).sum()
    }

    /// Cache tokens across all models (cache read + cache write).
    pub fn cache_tokens(&self) -> u64 {