Module usage 

Token-usage accounting summed across the whole optimization run, plus the one-line footer (prompt:24.3k tokens:515) printed at the bottom of the output. The footer shape mirrors boro’s progress UI.

prompt is the cumulative input-token count across every model request the run made (each tool-loop iteration resends the growing message history, so it grows fast); completion is the cumulative output tokens. cache_read / cache_creation are shown only when nonzero (prompt-cache hits/writes).

Structs

Usage

Functions

fmt_scaled 🔒

fmt_tokens 🔒

Compact token count: exact below 1000, else scaled to k/M/G with one decimal (a trailing .0 is dropped). Matches boro’s footer formatting.