[WRONG BRANCH] fix(codex): drop expired non-blocking quota evidence

[luvs01]

Motivation

• A retained elapsed short-window tuple (low percent) in identity-bound main-policy snapshots could take precedence over a fresh exhausted weekly reading and let requests continue using the stored main credential despite a blocking weekly change.
• The intent is to keep expired blocking evidence (that was already a hard-lock) but not to let stale non-blocking short readings override fresh blocking weekly evidence.

Description

• Only preserve elapsed short-window evidence when it is blocking by the hard-lock threshold, by centralizing MAIN_ACCOUNT_HARD_LOCK_PERCENT in src/codex/quota-types.ts and checking it during carry/merge.
• In mergeAccountQuota and assignCarriedShort (src/codex/quota.ts) change the retention rules so expired short tuples are dropped unless they represent a blocking reading (>= hard-lock percent).
• Make finitePercent a type-guard and avoid repeated optional lookups when checking existing short percent.
• Add a regression test that reproduces the reported sequence (elapsed shortPercent 4 followed by weeklyPercent 99) to assert the stale non-blocking short is discarded and the fresh weekly 99 blocks admission (tests/codex-integration/main-account-hard-lock-policy.test.ts).
• Update architecture documentation to clarify that expired blocking short evidence remains retained while expired non-blocking evidence is dropped (structure/providers/openai-tiers.md and related docs).

Testing

• Ran the focused regression suite with bun test tests/codex-integration/main-account-hard-lock-policy.test.ts and it passed (25 tests, 0 failures).
• Typecheck and repository gates passed with bun run typecheck, bun run structure:check, and bun run privacy:scan.
• Ran git diff --check and bun run test (full suite); the focused change passed, while the full parallel test run failed for unrelated, environment-sensitive test harness issues (process-group cleanup failures in other tests).
• The change includes the new regression test to prevent reintroduction of the stale non-blocking-short bypass.

---

Codex Task

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Advanced

Run ID: 195ea633-b245-4b65-9ea4-ff6a0c41f8e4

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

✅ Deterministic PR hygiene checks passed.

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-14T00:22:46.739345Z	caf03d0	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[github-actions]

⏳ DRAFT

• wrong target branch (Dev); retarget to dev. UI screenshot required.

What to do

• Retarget this PR to dev — all contributions go to dev.
• Add a screenshot of the UI change to the PR description.

Its title has been prefixed with [WRONG BRANCH].
This pull request is being kept as a draft automatically. Once every issue above is resolved, it will be marked ready for review again.