Repository navigation
refactor(bot): break Cycle C via direct imports + extract audioFeatures/vcWeights (#889) - #931
Merged
Merged
Conversation
…es/vcWeights Resolves #889 — last runtime circular dep in packages/bot/src. The cycle was queueManipulation.ts <-> autoplay/replenisher.ts via a 6-symbol barrel import. 4 of those symbols already lived in standalone files (candidateFallback.ts); the other 2 needed minimal extraction: - getTrackAudioFeatures + audioFeatureCache (LRUCache) -> autoplay/audioFeatures.ts - buildVcContributionWeights -> autoplay/vcWeights.ts replenisher.ts now imports each symbol directly from its source file. queueManipulation.ts and queueRescue.ts keep their public surface via re-export, so no consumer change. Madge runtime cycles: 2 -> 1 (deferred type-only types/CustomClient cycle remains, as documented in #871). Tests: 433/433 autoplay+queue+replenisher pass. 3075/3075 total bot suite pass (2 unrelated prom-client env failures in worktree only). The closure-capture failure mode from the prior agent attempt is avoided by keeping audioFeatureCache as a module-level const in its new file -- purely lexical relocation, no shape change. ADR: docs/decisions/2026-05-21-cycle-c-direct-imports-and-audio-features-extraction.md
Contributor
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
![greptile-apps[bot]](https://avatars.githubusercontent.com/in/867647?s=60&v=4){kind=small}
There was a problem hiding this comment.
Your free trial has ended. If you'd like to continue receiving code reviews, you can add a payment method here.
|
Warning Rate limit exceeded
You’ve run out of usage credits. Purchase more in the billing tab. ⌛ How to resolve this issue?After the wait time has elapsed, a review can be triggered using the We recommend that you space out your commits to avoid hitting the rate limit. 🚦 How do rate limits work?CodeRabbit enforces hourly rate limits for each developer per organization. Our paid plans have higher rate limits than the trial, open-source and free plans. In all cases, we re-allow further reviews after a brief timeout. Please see our FAQ for further information. ℹ️ Review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Pro Run ID: ⛔ Files ignored due to path filters (1)
📒 Files selected for processing (6)
✨ Finishing Touches🧪 Generate unit tests (beta)
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. Comment |
There was a problem hiding this comment.
Your free trial has ended. If you'd like to continue receiving code reviews, you can add a payment method here.
|
Failed to generate code suggestions for PR |
|
This was referenced May 22, 2026
Merged
LucasSantana-Dev
added a commit
that referenced
this pull request
May 22, 2026
…es/vcWeights (#889) (#931) Resolves #889 — last runtime circular dependency in \`packages/bot/src\`. ## What changed Cycle C was \`queueManipulation.ts ↔ autoplay/replenisher.ts\` via a 6-symbol barrel import. 4 of those symbols already lived in standalone files (\`candidateFallback.ts\`); the other 2 needed minimal extraction: - \`getTrackAudioFeatures\` + \`audioFeatureCache\` → \`autoplay/audioFeatures.ts\` - \`buildVcContributionWeights\` → \`autoplay/vcWeights.ts\` \`replenisher.ts\` now imports each symbol directly from its source. \`queueManipulation.ts\` and \`queueRescue.ts\` retain their public surface via re-export, so no consumer touches. ## Madge - Before: 2 cycles (Cycle C + deferred type-only \`types/CustomClient\` cycle) - After: 1 cycle (only the deferred type-only one remains, as documented in #871) Unblocks promotion of \`.github/workflows/madge.yml\` from \`continue-on-error: true\` to a hard gate (separate follow-up PR). ## Tests - 433/433 autoplay + queue + replenisher tests pass - 3075/3075 total bot suite pass (2 unrelated \`prom-client\` env failures in worktree only) ## Closure-capture avoidance The prior agent attempt at Cycle C broke 107 tests due to closure-captured state. This change keeps \`audioFeatureCache\` as a module-level \`const LRUCache\` — purely lexical relocation, no shape change. ## ADR \`docs/decisions/2026-05-21-cycle-c-direct-imports-and-audio-features-extraction.md\` records the decision, alternatives (B: larger extraction; C: accept the cycle; D: callback indirection), and revisit triggers. ## Changelog [Unreleased] Changed: break the last runtime circular dependency in \`packages/bot/src\` (Cycle C, #889).
LucasSantana-Dev
added a commit
that referenced
this pull request
May 22, 2026
10 PRs accumulated since v2.13.0 (cut 2026-05-21, shipped same-day). ## Added - feat(bot): autoplay closed-loop telemetry writers (Phase B) #933 - feat(backend): /recommendations/history read path (Phase C) #935 - feat(download): cover + re-enable /download command #930 - feat(music): cover + re-enable collaborative playlist #929 - feat(management): /customcommand coverage #928 - feat(management): cover + re-enable /embed builder #927 - test(bot/recommendation): cover 4 untested handlers #926 ## Changed - refactor(bot): Cycle C closure of #889 — runtime cycles 2 to 1 #931 - refactor(bot/autoplay): Cycles A + B residuals — 4 to 2 #925 - chore(prisma): Recommendation model repurposed for telemetry #932 Roadmap status: Phase A+B+C of the autoplay recommendation roadmap shipped in this release. Phase D (session-coherence layer) gated on 7 days of production baseline -- revisit 2026-05-29. Also bundled via release-branch-autosync: fix(ci): v-prefix trivy-action tag (#934) -- unblocks production Docker image republishing.
LucasSantana-Dev
added a commit
that referenced
this pull request
May 22, 2026
v2.14.0 — 10 PRs since v2.13.0 (cut 2026-05-21). ## Highlights **Autoplay recommendation telemetry (3-phase roadmap, ADR \`2026-05-21-autoplay-recommendation-roadmap\`):** - Phase A (#932) — Prisma \`Recommendation\` model repurposed for closed-loop autoplay telemetry. New \`RecommendationSource\` enum, \`signals: Jsonb\`, \`discordUserId\`, aggregation index. - Phase B (#933) — non-throwing writers. Every autoplay pick writes a row; \`playerFinish\` and \`playerSkip\` flip the outcome flags using 30% / 5s thresholds (exported constants for tuning). - Phase C (#935) — read path. \`GET /api/guilds/:guildId/recommendations/history?days=<n>\` returns per-source acceptance rate + global summary aggregated over the window (default 7d, clamped [1, 30]). - **Phase D deferred.** Gated on 7 days of production baseline starting after this release deploys. If per-source acceptance \u2265 85%, Phase D becomes optional. Revisit 2026-05-29. **bot/autoplay circular-dependency closure (#889 done):** - #925 — Cycles A+B (queueMarkers, candidateContracts) — madge 4 → 2. - #931 — Cycle C (audioFeatures, vcWeights) — madge 2 → 1 (only the deferred type-only \`types/CustomClient\` cycle remains). Unblocks promoting \`madge.yml\` to a blocking CI gate. **Feature toggle re-enables (each with full test coverage):** - \`/download\` (#930), collaborative playlist (#929), \`/embed\` builder (#927). - Test-only coverage adds: \`/customcommand\` (#928), 4 recommendation handlers (#926). **Bundled via release-branch-autosync:** \`fix(ci): v-prefix trivy-action tag\` (#934) — unblocks production Docker image republishing on main. ## Verification - All 10 PRs passed CI (Quality Gates, SonarCloud, CodeQL, semgrep, knip, lint, madge, security) before merging to release/v2.14.0. - Bot suite: 3109+ tests passing across the autoplay + queue + replenisher + recommendation + trackHandlers surface. - Backend suite: 343+ integration tests passing including the new \`/recommendations/history\` route. - Shared suite: 12 new tests on \`recommendationTelemetryReadService\`. ## Deploy checklist After merge: 1. Tag \`v2.14.0\` on the merge commit. 2. Verify the Docker publish workflow succeeds on the merge commit (was the original failure mode pre-#934; now fixed). 3. Confirm homelab pulls the new images. 4. Begin 7-day baseline collection for the autoplay recommendation telemetry. ## Migrations One Prisma migration in this release: \`20260521000000_recommendation_telemetry_phase_a\`. Repurposes the unused \`recommendations\` table: - Drops \`algorithm: String\` (was scaffolded, never written by any service). - Adds \`recommendationSource\` enum + \`signals\` Jsonb + \`discordUserId\` + indexed by \`(guildId, source, createdAt)\`. - \`confidence\` made nullable. Safe to deploy: no production rows exist in the table prior to this release. <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added recommendation history endpoint to retrieve past recommendations and acceptance rates by source * Telemetry recording now tracks recommendation outcomes (accepted/rejected) based on listening behavior * New feature toggles for Embed Builder and Collaborative Playlist support * **Chores** * Bumped all package versions to 2.14.0 * Updated database schema to support recommendation source tracking and telemetry * **Tests** * Added comprehensive test coverage for download and autoplay features, collaborative playlist, and recommendation handlers <!-- review_stack_entry_start --> [](https://app.coderabbit.ai/change-stack/LucasSantana-Dev/Lucky/pull/936?utm_source=github_walkthrough&utm_medium=github&utm_campaign=change_stack) <!-- review_stack_entry_end --> <!-- end of auto-generated comment: release notes by coderabbit.ai -->
](https://app.coderabbit.ai/change-stack/LucasSantana-Dev/Lucky/pull/936?utm_source=github_walkthrough&utm_medium=github&utm_campaign=change_stack){kind=link}
This branch was successfully deployed
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.
Resolves #889 — last runtime circular dependency in `packages/bot/src`.
What changed
Cycle C was `queueManipulation.ts ↔ autoplay/replenisher.ts` via a 6-symbol barrel import. 4 of those symbols already lived in standalone files (`candidateFallback.ts`); the other 2 needed minimal extraction:
`replenisher.ts` now imports each symbol directly from its source. `queueManipulation.ts` and `queueRescue.ts` retain their public surface via re-export, so no consumer touches.
Madge
Unblocks promotion of `.github/workflows/madge.yml` from `continue-on-error: true` to a hard gate (separate follow-up PR).
Tests
Closure-capture avoidance
The prior agent attempt at Cycle C broke 107 tests due to closure-captured state. This change keeps `audioFeatureCache` as a module-level `const LRUCache` — purely lexical relocation, no shape change.
ADR
`docs/decisions/2026-05-21-cycle-c-direct-imports-and-audio-features-extraction.md` records the decision, alternatives (B: larger extraction; C: accept the cycle; D: callback indirection), and revisit triggers.
Changelog
[Unreleased] Changed: break the last runtime circular dependency in `packages/bot/src` (Cycle C, #889).