Module usage_processor 

Per-provider usage accumulation: merges the token fields of one session record into a running per-model map, normalising provider-specific shapes along the way.

Structs

CodexTokenTotals

Snapshot of Codex’s cumulative total_token_usage counters.

Constants

CODEX_TOKEN_FIELDS

The five integer fields of a Codex total_token_usage object.

Functions

accumulate_i64_fields

Adds the named i64 fields from source into target, in place.

accumulate_nested_object

Adds every i64 field of source_nested into the nested object stored at target_obj[field_name].

claude_request_context

The full prompt context of one Claude request: non-cached input plus cache reads plus cache writes — what the provider compares against the “above Nk tokens” threshold.

process_claude_usage

Merges one Claude usage record into conversation_usage, keyed by model.

process_codex_usage

Merges one Codex usage delta into conversation_usage, keyed by model.

process_gemini_usage

Merges one Gemini usage record into conversation_usage, keyed by model.