Feature request: utilize codex reset from codexbar

[jtl06]

Hi, OpenAI recently added the ability to reset Codex limits on demand. It would be nice If I could reset my limits using Codexbar, rather than needing to use the codex app. Not sure how this could be implemented, but if it is possible, it would be a very nice UX improvement.

Thanks

[clawsweeper]

Codex review: keeping this open for maintainer follow-up; there is still a little grit to resolve. Reviewed June 13, 2026, 9:54 PM ET / 01:54 UTC.

Summary
Keep open: current main exposes Codex usage, reset countdowns, dashboard extras, and a Buy Credits flow, but it does not implement the banked reset-credit endpoints or any in-app reset-limits action. The related external codex-reset note is useful implementation context, but this still needs maintainer product/auth sign-off before code work is safe.

Reproducibility: not applicable. this is a feature request for a new Codex action, not a report of broken current behavior. Source inspection confirms current main lacks the requested reset-credit surface.

Ways to help us reproduce this

• Add a screenshot or short recording showing the behavior.
• Add expected vs actual behavior.
• Include redacted logs or terminal output.

Next step
A maintainer should first decide whether CodexBar should perform provider-side reset-credit redemption and what confirmation/account-safety UX is required.

Review details

Best possible solution:

Design a maintainer-approved Codex action that lists available reset credits for the active account, requires explicit user confirmation, redeems through a stable OpenAI flow, then refreshes usage state without weakening account routing or cookie/OAuth privacy boundaries.

Do we have a high-confidence way to reproduce the issue?

Not applicable: this is a feature request for a new Codex action, not a report of broken current behavior. Source inspection confirms current main lacks the requested reset-credit surface.

Is this the best way to solve the issue?

Unclear until maintainer sign-off: the linked external tool gives a plausible endpoint shape, but the best CodexBar implementation needs explicit confirmation, account scoping, and provider auth/privacy approval before code work starts.

AGENTS.md: found and applied where relevant.

Remaining risk / open question:

• The reset-credit endpoints are undocumented and include a POST that consumes a finite account credit, so CodexBar integration needs explicit confirmation UX and provider auth/privacy review.
• The adjacent rolling-window auto-start PR is open and proof-blocked, so it should not be treated as a merged or safe superseding implementation.
• A naive implementation could redeem against the wrong active Codex account unless it reuses the existing account-routing and dashboard ownership safeguards.

Codex review notes: model internal, reasoning high; reviewed against 15c70c7d90fc.

Label changes

Label changes:

• add P3: This is a low-urgency UX feature request that needs product/auth review rather than urgent bug repair.
• add impact:auth-provider: The requested reset action would use Codex/OpenAI account auth and consume provider-side reset credits for the active account.
• add issue-rating: 🌊 off-meta tidepool: Current issue advisory state selects this label.
• add clawsweeper:no-new-fix-pr: Current issue advisory state selects this label.
• add clawsweeper:needs-maintainer-review: Current issue advisory state selects this label.
• add clawsweeper:needs-product-decision: Current issue advisory state selects this label.

Label justifications:

• P3: This is a low-urgency UX feature request that needs product/auth review rather than urgent bug repair.
• impact:auth-provider: The requested reset action would use Codex/OpenAI account auth and consume provider-side reset credits for the active account.

Evidence reviewed

What I checked:

• Issue request is still open: The issue body requests an in-app Codex limit reset action; it has no labels, no closing PR references, and only the ClawSweeper placeholder comment.
• External reset-credit context: Related issue Thanks — extended your wham notes into a Linux CLI redeem tool (codex-reset) #1485 links https://github.com/aaamosh/codex-reset and identifies GET /backend-api/wham/rate-limit-reset-credits plus POST /backend-api/wham/rate-limit-reset-credits/consume, but that item is informational rather than a CodexBar implementation.
• Codex provider docs lack reset redemption: docs/codex.md documents OAuth usage, OpenAI web dashboard extras, CLI RPC, reset timestamps, and credits purchase URL scraping, but no reset-credit listing or consume flow. (docs/codex.md:10, 15c70c7d90fc)
• Current OAuth path only fetches usage: CodexOAuthUsageFetcher defines /wham/usage and sends a GET with bearer auth; no reset-credit POST path is present there. (Sources/CodexBarCore/Providers/Codex/CodexOAuth/CodexOAuthUsageFetcher.swift:232, 15c70c7d90fc)
• Current dashboard path only fetches usage/dashboard data: OpenAIDashboardFetcher targets the Codex analytics page and /backend-api/wham/usage; the reset-credit endpoints from the related note are absent from current main. (Sources/CodexBarCore/OpenAIWeb/OpenAIDashboardFetcher.swift:22, 15c70c7d90fc)
• Codex provider menu actions lack reset action: CodexProviderImplementation adds credits display, login, and a System Account submenu, but no reset-limits action entry. (Sources/CodexBar/Providers/Codex/CodexProviderImplementation.swift:197, 15c70c7d90fc)

Likely related people:

• steipete: Current Codex OAuth, OpenAI dashboard, provider implementation, and docs lines blame to 7717813, and symbol history shows Peter introduced the Codex OAuth usage fetch and much of the OpenAI web dashboard surface. (role: current Codex/OpenAI surface owner; confidence: high; commits: 7717813b9eed, fca878568eda, 8dc9b333fa88; files: Sources/CodexBarCore/Providers/Codex/CodexOAuth/CodexOAuthUsageFetcher.swift, Sources/CodexBarCore/OpenAIWeb/OpenAIDashboardFetcher.swift, Sources/CodexBar/Providers/Codex/CodexProviderImplementation.swift)
• ratulsarna: Symbol history shows multiple OpenAI dashboard, battery, retry, and multi-account changes adjacent to the dashboard/auth surface a reset action would reuse. (role: historical OpenAI dashboard contributor; confidence: medium; commits: dd200dac317d, 2bea4f045b65, 0199e1ca9a00; files: Sources/CodexBarCore/OpenAIWeb/OpenAIDashboardFetcher.swift)
• LeoLin990405: Symbol history for additional_rate_limits points to the Codex Spark WHAM work, which is adjacent to reset-credit data from WHAM endpoints. (role: recent adjacent Codex WHAM contributor; confidence: medium; commits: afe4e802239f; files: Sources/CodexBarCore/Providers/Codex/CodexAdditionalRateLimitMapper.swift, Sources/CodexBarCore/OpenAIWeb/OpenAIDashboardFetcher.swift, docs/codex.md)
• Quicksaver: Open PR Add opt-in rolling window auto-start #1481 explores opt-in rolling-window auto-start through provider CLI prompts, which is adjacent to quota reset behavior but not a replacement for this request. (role: adjacent open PR author; confidence: low; commits: cf9011dbff5a, 9083c6796a55, 34a2eba032db; files: Sources/CodexBar/RollingWindowAutoStart.swift, Sources/CodexBar/UsageStore+RollingWindowAutoStart.swift, Tests/CodexBarTests/RollingWindowAutoStartTests.swift)

How this review workflow works

• ClawSweeper keeps one durable marker-backed review comment per issue or PR.
• Re-runs edit this comment so the latest verdict, findings, and automation markers stay together instead of adding duplicate bot comments.
• A fresh review can be triggered by eligible @clawsweeper re-review comments, exact-item GitHub events, scheduled/background review runs, or manual workflow dispatch.
• PR/issue authors and users with repository write access can comment @clawsweeper re-review or @clawsweeper re-run on an open PR or issue to request a fresh review only.
• Maintainers can also comment @clawsweeper review to request a fresh review only.
• Fresh-review commands do not start repair, autofix, rebase, CI repair, or automerge.
• Maintainer-only repair and merge flows require explicit commands such as @clawsweeper autofix, @clawsweeper automerge, @clawsweeper fix ci, or @clawsweeper address review.
• Maintainers can comment @clawsweeper explain to ask for more context, or @clawsweeper stop to stop active automation.

[aaamosh]

For anyone implementing this in CodexBar — the endpoint contract is small enough to fit in a comment. Verified live against a real Plus account on 2026-06-13 (we sent the actual POST and burned a real banked credit; weekly window went 100% → 0%, windows_reset: 1 in the response).

Base: https://chatgpt.com/backend-api. Same Authorization: Bearer <access_token> + ChatGPT-Account-Id: <account_id> headers CodexBar already uses for /wham/usage.

GET  /wham/rate-limit-reset-credits           → { credits: [{id, reset_type:"codex_rate_limits", status:"available"|"redeemed", granted_at, expires_at, ...}], available_count }
POST /wham/rate-limit-reset-credits/consume   body {"credit_id": "<id>", "redeem_request_id": "<uuid4>"}
                                              → { code: "reset", credit:{status:"redeemed", redeemed_at}, windows_reset: 1 }

Edge case worth handling: in the consume response the server collapses title, description, profile_image_url, profile_user_id to null even though those fields are populated in the GET /credits listing. Be lenient on those in the response model.

Endpoints came from the official openai.chatgpt VS Code extension's webview bundle (webview/assets/codex-api-*.js). There's a tiny no-deps Python reference implementation at https://github.com/aaamosh/codex-reset if it helps as a sanity check — it's the same three calls plus a before/after pair on /wham/usage to demonstrate the reset took effect.

Soju06/codex-lb#1015 was an attempt at the same feature; it's currently closed pending rework but the consume client there (code) matched the payload exactly, so the contract above should be portable to whatever CodexBar wants to do with it.

[steipete]

#1690 now contains the repaired, mechanically reviewed implementation at exact head 00da631f5. A request stub verifies the consume path, selected-account header, credit ID, and unique request ID without performing a live provider-side write; focused suites, localization coverage, make check, packaged-app proof, and autoreview are green.

The remaining question is product/provider-contract authority, not code completeness: should CodexBar spend a finite reset credit through an authenticated private POST? My recommendation is do not land or close this issue yet; keep CodexBar read-only until that write contract is documented or explicitly approved. The PR body states the exact decision and tradeoff.