Separate provider polling from passive CLI file refresh - #142
Conversation
…urce-refresh # Conflicts: # docs/EFFORT-LOG.md
This comment has been minimized.
This comment has been minimized.
|
You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard. |
|
@kody review |
🤔 Insufficient Task ContextI found a task linked to this PR, but it only contains minimal information (title only, no description or acceptance criteria). To perform a meaningful business rules validation, I need more details. 🔍 What I need to validate:
💡 How to improve the task context:
|
…dy review on provider issue clearing
|
@kody please review |
…file windows Do not wipe provider issues when reading cached local windows. Only clear OpenAI issues when a fresh session file window is present.
This comment has been minimized.
This comment has been minimized.
The Independent Provider and File Refresh board row lived only in docs/EFFORT-LOG.md while root EFFORT-LOG.md stayed stale. Add the same PR #142 row to both ledgers with current implementation and CI-based verification wording. Co-authored-by: Jay Wedgeworth <jaywedgeworth22@users.noreply.github.com>
This comment has been minimized.
This comment has been minimized.
UsageHistoryView init ran on every PlatformDetailPage body evaluation, including the 30-second clock tick, and called historySamples() each time. Drop the custom initializer so samples stay empty until reload() runs from onAppear and lastChecked changes; MonitorModel already memoizes file parses. Co-authored-by: Jay Wedgeworth <jaywedgeworth22@users.noreply.github.com>
|
Designer UI pass on tip fe4b872. APPROVE. Settings gained Provider Checks, Provider Check Interval, Codex Session File Checks, and Session File Check Interval. The pickers say Every 1 minute and Every N minutes. Explanatory lines use sentenceGap, a real NBSP then a space. Turning provider checks off gates the refresh. Usage history still loads when the view appears. RECENT USAGE HISTORY matches the other eyebrows. Extra-ship no. Designer is not merging. |
State explicitly that the notification screenshot informed design only, not Mac UI acceptance. Match the Validation wording Kody requested for code review plus CI fixtures and iOS simulator screenshots. Co-authored-by: Jay Wedgeworth <jaywedgeworth22@users.noreply.github.com>
This comment has been minimized.
This comment has been minimized.
|
Sentence-spacing violation in the PR description leaves only one space between “independently.” and “It,” after “paths.” before “Scheduled,” and after “work.” before “Disabling.” Add a second space after every sentence-ending period to meet the two-literal-space rule. Kody rule violation: Use two spaces between sentences in every human-facing string and agent-written paragraph |
|
Designer UI pass on tip 008b5c7. QUIET. The new commit only edits docs/design/platform-usage-history.md. It rewords the review-evidence paragraph and the Validation section. No product screen, control, label, or user-facing string changes. Sentence spacing in the new paragraphs holds: two spaces after "PR #134.", "window.", "Mac UI.", and "tests." Existing Title Case headers are untouched. No new buttons, menus, or prices. The prior Designer pass on fe4b872 was APPROVE and listed no blockers. This diff does not reopen Provider Checks, the interval pickers, or usage-history loading. |
|
@kody review |
Superseded by fixes and passing CI
…cord Add explicit init(model:state:row:) to UsageHistoryView to preserve internal accessibility across files without early computing samples in view init. Add repo-first pre-work claim record to EFFORT-LOG.md and docs/EFFORT-LOG.md.
Kody Review CompleteGreat news! 🎉 Keep up the excellent work! 🚀
|
| Options | Enabled |
|---|---|
| Bug | ✅ |
| Performance | ✅ |
| Security | ✅ |
| Business Logic | ✅ |
Depends on #139. Refs #137 and #136.
Summary
This PR separates active provider quota polling from passive Codex session-file monitoring so each source can run, persist, and be disabled independently. It also adds local quota history charts, measured runaway-usage comparisons, and alert-to-platform navigation.
Functional changes
Independent quota refresh paths
Passive Codex session reading
token_countquota events are admitted; unrelated events and quota pools do not produce readings.Local usage history and runaway comparisons
Console and notification navigation
Managed settings and validation
PULL_ENDPOINT,PUSH_ENDPOINT, andSETTINGS_REFRESH_SECONDSby name instead of listing all root secrets.Outstanding automated review findings
The provided patch has one critical and five high findings:
UsageHistoryViewcan load and parse the complete sample file during initialization on every SwiftUI re-render.docs/EFFORT-LOG.md.