Adds model availability, clearer quota reporting and an online report submenu to Codex Usage.
Added
- Added model availability to online reports, with GPT-6 Astra always visible and distinct statuses for unavailable, missing and conflicting data.
- Added used and remaining percentages, reported periods and reset times for main and additional quotas. Alternative models appear only when explicitly reported for a blocked model.
- Added an online submenu for model availability, main limits, additional limits, daily usage, credits, profile statistics and per-endpoint technical details.
Changed
- Online submenu views share one snapshot, fetched when a view is first selected. Navigation reuses it; refresh replaces all online data and shows the query time and unavailable endpoints.
- Explanatory prose now uses the terminal’s natural wrapping. Long model and quota values use wrapped records when tables would truncate them.
- CSV exports prioritise model availability, alternative-model and additional-limit fields within the existing row limit.
Fixed
- The interactive menu stays open when the Codex login is missing or unreadable, so local reports and refresh remain available.
- Quota periods follow the reported duration instead of assuming a weekly secondary window. Invalid or missing percentages appear as unknown values.
- Interactive reports preserve the selected colour mode, including
--no-colour.
Compatibility and network behaviour
- Existing commands and JSON/CSV structures remain supported. Direct online reports and TXT exports still include the complete report without interactive prompts.
- Model availability uses the existing read-only usage response, with no extra requests per model. The tool does not call or switch models, buy credits or redeem resets.
- Availability and quota percentages are not an exact remaining message count for Astra.