perf(spend): parallelize provider baselines in makeRequest - #3099
Closed
Yuxin-Qiao wants to merge 1 commit into
Closed
perf(spend): parallelize provider baselines in makeRequest#3099Yuxin-Qiao wants to merge 1 commit into
Yuxin-Qiao wants to merge 1 commit into
Conversation
Refresh independent spend snapshots concurrently via TaskGroup instead of sequential await, cutting 400ms-6s additive latency when multiple providers (Claude/Cursor/OpenCodex) share the 365-day window.
|
🦞👀 Pull request received. I will update this pull request when review starts. |
clawsweeper
Bot
added
P2
|
Codex review: needs real behavior proof before merge. Reviewed August 20, 2026, 6:08 AM ET / 10:08 UTC. ClawSweeper reviewWhat this changesThe PR refreshes eligible non-Codex provider spend snapshots concurrently before building the combined spend-dashboard request. Merge readiness⛔ Blocked until real behavior proof is added - 3 items remain Keep open: source review finds the fully drained task group consistent with the dashboard’s refresh-and-recapture contract, but required after-fix real behavior proof is missing. Priority: P2 Review scores
Verification
Live VerificationCommand: Result: FAIL (failed) — execution before step 1 Assertions:
How this fits togetherCodexBar’s spend dashboard refreshes provider-specific cost snapshots before it captures inputs for the dashboard model. Those inputs then drive the menu and dashboard’s combined usage-and-spend display. flowchart LR
A[Dashboard refresh request] --> B[Provider baselines]
B --> C{Refresh required?}
C -->|Yes| D[Concurrent provider refreshes]
C -->|No| E[Current provider publications]
D --> F[Capture refreshed publications]
E --> F
F --> G[Combined spend dashboard]
G --> H[Menu and dashboard display]
Before merge
Agent review detailsSecurityNone. Review metrics
Merge-risk optionsMaintainer options:
Technical reviewBest possible solution: Keep the fully drained task group and add redacted after-fix evidence from a real multi-provider dashboard refresh before merge. Do we have a high-confidence way to reproduce the issue? Not applicable: this is a performance optimization, not a separately reported malfunction with a failing reproduction. Is this the best way to solve the issue? Yes: a drained task group is the narrowest maintainable way to overlap independent provider refreshes while retaining the existing post-refresh capture boundary. AGENTS.md: found and applied where relevant. Codex review notes: model internal, reasoning high; reviewed against a3457dafbe63. LabelsLabel changes:
Label justifications:
EvidenceWhat I checked:
Likely related people:
Rank-up movesOptional improvements that raise the rating; they are not merge blockers.
Rating scale
Overall follows the weaker of proof and patch quality. Workflow
|
Contributor
Author
|
Superseded by larger perf PR (parallel+memoize) per feedback |
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
1 participant
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.
Parallelize independent spend snapshot refreshes in
SpendDashboardSource.makeRequest:237.Before:
for baseline where shouldRefresh { await store.refreshProvider / refreshSpendDashboardTokenUsageNow }serially awaited per provider. With 3 independent providers (Claude/Cursor/OpenCodex) each 400ms-3s (Cursor pagination), total 1.2-9s additive.After:
withTaskGroupconcurrent, wall-time = slowest provider only. Cold全部first-paint from ~3-10s → ~2-3s (Codex scan still dominates).Scope: single function, no new persistence, no behavior change beyond concurrency.
SpendDashboardSourceConcurrencyTestscovers same-scope ownership.Fixes P0-4 from spend audit.
Verified:
swiftformat+swiftlint --strictclean.