Refresh web console controls and Codex quota display - #983
Closed
MapleStoryIdle wants to merge 1 commit into
Closed
Conversation
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?s=60&v=4){kind=small}
There was a problem hiding this comment.
Findings
- [Major] Refetch Codex limits when the selected model changes —
getCodexSubscriptionLimitsnow selects a model-specific limit bucket, but the web query is keyed only bysessionIdand is infinitely stale. After switching between a model with its own bucket and a regular Codex model, the header badge can keep showing the old model's quota until a later thinking transition happens. Evidence:web/src/hooks/queries/useCodexSubscriptionLimits.ts:22, related selector contextcli/src/modules/common/codexSubscriptionLimits.ts:76.
Suggested fix:// SessionHeader.tsx const codexLimitsState = useCodexSubscriptionLimits({ api, sessionId: session.id, model: session.model ?? null, enabled: session.active && session.metadata?.flavor === 'codex', thinking: props.status?.thinking ?? session.thinking }) // useCodexSubscriptionLimits.ts queryKey: [ ...queryKeys.sessionCodexSubscriptionLimits(sessionId ?? 'unknown'), args.model ?? null ],
Summary
- Review mode: initial
- One issue found: Codex quota display can become stale after model changes because the client cache key does not include the model. Residual risk: web UI paths are not covered by automated tests in this run.
Testing
- Not run (automation:
bunis not installed in this runner)
HAPI Bot
| const thinking = args.thinking === true | ||
| const enabled = Boolean(args.enabled && api && sessionId) | ||
| const query = useQuery({ | ||
| queryKey: queryKeys.sessionCodexSubscriptionLimits(sessionId ?? 'unknown'), |
There was a problem hiding this comment.
[MAJOR] This query needs to vary with the selected model. The CLI selector now chooses a model-specific bucket when model matches limitId/limitName, but this cache key only contains sessionId and staleTime is infinite. If the user switches between Spark and a regular Codex model, the header badge can keep rendering the previous model's quota until a later thinking transition refetches it.
Suggested fix:
// include model in the hook args from SessionHeader
model: session.model ?? null,
// and key by it here
queryKey: [
...queryKeys.sessionCodexSubscriptionLimits(sessionId ?? 'unknown'),
args.model ?? null
],
MapleStoryIdle
force-pushed
the
feat/codex-desktop-web-ui
branch
from
July 1, 2026 03:01
a0a37fd to
b304743
Compare
There was a problem hiding this comment.
Findings
- [Minor] Avoid fetching hidden Telegram-only header data —
SessionHeaderreturnsnullin Telegram, but the new Codex limits hook runs before that return and is enabled for every active Codex session. Opening a Codex session in the Mini App now triggers/codex-subscription-limitsand an RPC that can spawncodex app-server, even though the badge is never rendered. Evidence:web/src/components/SessionHeader.tsx:211,web/src/components/SessionHeader.tsx:246.
Suggested fix:const telegramHeaderHidden = isTelegramApp() const codexLimitsState = useCodexSubscriptionLimits({ api, sessionId: session.id, model: session.model ?? null, enabled: !telegramHeaderHidden && session.active && session.metadata?.flavor === 'codex', thinking: props.status?.thinking ?? session.thinking }) if (telegramHeaderHidden) { return null }
Summary
- Review mode: follow-up after new commits
- Prior stale-model-cache issue no longer reproduces because the query key now includes
model. One minor issue remains: Telegram opens can still trigger the new quota RPC for a header that is not displayed. Residual risk: this review did not run the full browser UI manually.
Testing
- Not run (automation)
HAPI Bot
| session.id, | ||
| session.metadata?.flavor ?? null | ||
| ) | ||
| const codexLimitsState = useCodexSubscriptionLimits({ |
There was a problem hiding this comment.
[MINOR] This hook still runs before the Telegram early-return below, so active Codex sessions opened in the Mini App will fetch /codex-subscription-limits and route an RPC that can spawn codex app-server, even though SessionHeader returns null and never renders the quota badge in Telegram.
Suggested fix:
const telegramHeaderHidden = isTelegramApp()
const codexLimitsState = useCodexSubscriptionLimits({
api,
sessionId: session.id,
model: session.model ?? null,
enabled: !telegramHeaderHidden && session.active && session.metadata?.flavor === 'codex',
thinking: props.status?.thinking ?? session.thinking
})
if (telegramHeaderHidden) {
return null
}
Owner
|
Closing as part of repository maintenance: this PR has merge conflicts with the main branch and has had no activity for over two weeks. If the change is still relevant, please rebase onto the latest main and feel free to reopen (or open a fresh PR). Thanks for the contribution! |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Testing
AI assistance