fix(voice): distinguish carrying and evidence exports - #780
Conversation
|
Note Reviews pausedIt looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the Use the following commands to manage reviews:
Use the checkboxes below for quick actions:
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (1)
🚧 Files skipped from review as they are similar to previous changes (1)
Included review availability: Your plan provides up to 1 included review per hour; 0 remain after this review. 📝 WalkthroughWalkthroughVoice 할당 CSV가 ChangesVoice 조합 정합성
Estimated code review effort: 2 (Simple) | ~10 minutes Merge Risk: ⚪ Minimal · up to The change adds explicit carrying and derivation-evidence CSV fields while preserving compatibility and updating related authority documentation and metadata; no actionable merge-blocking risk remains beyond normal checks and review. 🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
Full details: Docstring CoverageExplanation Docstring coverage is 66.67% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 3 functions across 3 files. (1 skipped: 1 unsupported.)
✨ 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 |
|
You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard. |
…ority-20260828 # Conflicts: # docs/product-technical-gap-baseline.md # uv.lock
|
You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard. |
|
You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard. |
Signed-off-by: Codex <codex@localhost>
Signed-off-by: Codex <codex@localhost>
Signed-off-by: Codex <codex@localhost>
Signed-off-by: Codex <codex@localhost>
Signed-off-by: Codex <codex@localhost>
Signed-off-by: Codex <codex@localhost>
|
Cycle 2026-09-03: independent APPROVE still required on exact head |
|
Cycle 2026-09-03T05:01Z: writes work. Independent APPROVE still required on the exact head. Copilot/Devin/CodeRabbit are not independent. No self-approve. Do not squash-merge without independent APPROVE + exact-head required success. Not mixing i18n #922 into this PR. |
Adopt #924's restored docs-contract test triggering while preserving Voice export authority delta without force rewrite.
Adopt protected #931 workflow/contract blobs on the Voice export branch while preserving the evidence/export product delta. This non-force two-parent convergence keeps the child stack repairable without transferring predecessor evidence.
seonghobae
left a comment
There was a problem hiding this comment.
Fresh parent/base review found #780 still anchored on the pre-#931 protected line. Its Voice/export product delta does not need to own or fork repository-local CI semantics. Repaired non-destructively at 1d8fa267b059289e77301a09985dfac70a439814 by adopting the exact protected #931 workflow/contract blobs and recording main@83eba56149eb802cd63642c507c324c9976ec78e as the second parent. Immediately restacked child #901 onto this moved parent at 9e582e4be080ae74bffac83817e639993ac85a71, preserving the child's exact baseline blob. No force rewrite or predecessor evidence transfer.
seonghobae
left a comment
There was a problem hiding this comment.
Fresh governance finding on the unchanged exact head: repository-local Ready-trigger Tests/Ontology are terminal GREEN, but the organization-ruleset required central workflow evidence on this SHA is still only the prior Draft-era cancelled Security/SAST/central-CodeQL/scheduler set; ready_for_review did not recreate those required-workflow runs. The separate GitHub Advanced Security CodeQL check is success but is not a substitute for the ruleset-pinned .github/workflows/codeql-pr.yml. Exact canary evidence has been handed to canonical owner ContextualWisdomLab/.github#712 (comment 5552270732). Keep #780 Ready as the stack's validation lane, but do not merge/bypass until the central owner recreates terminal required-workflow evidence on this unchanged head and a qualifying independent APPROVE exists.
seonghobae
left a comment
There was a problem hiding this comment.
Fresh control-plane correction: the body still describes .github#1927 as an unresolved single-entry dispatcher allowlist defect, but canonical owner evidence now shows the dual-principal value github-actions[bot],opencode-agent[bot] is live and has authorized the App sender. That does not manufacture the central Ready-transition executions absent from this exact head. Preserve 1d8fa267... unchanged; do not Draft/Ready churn or no-op commit to trigger checks. Missing Ready-state Security/SAST/CodeQL receipts and qualifying independent approval remain current evidence gaps, while .github#1902 still owns CodeQL exact-job recovery.
seonghobae
left a comment
There was a problem hiding this comment.
Fresh owner-path revalidation after .github#1927 authorization convergence: exact unchanged head 1d8fa267b059289e77301a09985dfac70a439814 remains mechanically mergeable and all existing review threads are resolved, but no qualifying independent APPROVE exists. Re-ran the exact-head central evidence instead of churning the branch: CodeQL PR run 33903233347 is now attempt 3; Security Scan 33903233493, SAST Semgrep 33903233503, and Required Noema Review 33903231974 are now attempt 2. All four were accepted by GitHub and entered queued state on this exact head. Do not count them GREEN until terminal receipts exist; do not merge or bypass while they are queued or while independent approval is absent.
|
Scheduled review-feedback autofix for this PR head.
|
Summary
Repairs Voice-of-X export authority. Exact-value CSV distinguishes the carrying Post from derivation-evidence Post, ADR 0251/0256 ownership is reconciled, and historical frontend/browser observations remain predecessor evidence after branch movement.
Protected-main convergence
Review
5116306008found this branch anchored on the pre-#931 protected line. Commit1d8fa267b059289e77301a09985dfac70a439814adopted the exact protected #931 workflow/contract blobs and joined the repaired product tree to protectedmain@83eba56149eb802cd63642c507c324c9976ec78ewithout rewriting the Voice/export delta.Exact live boundary
main@83eba56149eb802cd63642c507c324c9976ec78e1d8fa267b059289e77301a09985dfac70a439814APPROVEDreview; normal merge remains unavailableCurrent exact-head evidence
33952504686: success; Frontend101269793625and Full suite101269793798passed33952504679: success; ontology validation101269793684passed and publish correctly skipped on PR33903233493, attempt 2: success on the unchanged exact head; scope detection, Scorecard, and Trivy passed, while inapplicable jobs skipped33903233503, attempt 2: success on the unchanged exact head; exact submitted revision was verified, SARIF upload passed, and the gate found no Medium+ failure33903233347, attempt 3: failure. All three language compatibility jobs failed atRead current-head CodeQL dispatch verdictbecause no authenticated terminalcodeql-dispatch/*receipt exists; the dispatch job was skipped on rerun. This reproduces the canonical rerun-recovery defect owned byContextualWisdomLab/.github#1902. Consumer evidence is already posted there; do not issue further manual reruns, synthetic statuses, or head churn. The owner lane has since added structural pagination RED86898d3ecccdf8306d8dc42c8f9e7d5ee8dfbc3aand source repairdf35cfe57b90bfcf6caac1440390c057ddc48347, which paginates all five direct producer job/artifact collection pairs. Current-head review5137591346found the remaining acceptance gap: those tests assert shell source text rather than reproducing a unique exact job/artifact that exists only on page 2. Shard, coordinator, and settlement still need behavioral page-2 fixtures plus hosted exact-head GREEN and independent review before canonical promotion. The repair is neither protected nor immutably released, so fix(voice): distinguish carrying and evidence exports #780 remains blocked on owner completion and subsequent fresh consumer evidence.33903231974, attempt 2: failure, not pending. Exact-head admission, trusted workflow materialization, GitHub App credential mint, live-head validation, target-visibility resolution, and contextual-orchestrator sidecar provisioning all succeeded. The model phase then failed after 319.3s withHTTP 429 Too Many Requests,served_model=deepseek-ai/deepseek-v4-flash-0731, andcaller attempts=1; the workflow explicitly records that gateway owns repair/failover, and no verdict envelope was published. The sidecar was pinned to contextual-orchestrator414f22973658c4ddc3d4320fcf7acd9b4e8ba991. This is canonical-owner evidence forcontextual-orchestrator#1094, whose current contract keepsorchestrator/freestructuredresponse_formatrequests on the Fugu/TRINITY–Conductor control plane and re-selects retryable 429/502/timeout failures. Fresh fix(voice): distinguish carrying and evidence exports #780 consumer evidence was posted to #1094 as comment5579640679; no LineageWeave retry loop/provider pin is authorized.The earlier
.github#1927dispatcher-identity allowlist defect is no longer a live blocker: canonical owner evidence shows the dual-principal authorization is deployed. It does not manufacture a missing terminal CodeQL receipt;.github#1902remains the CodeQL owner path through protected promotion and consumer revalidation.Direct descendants remain stacked on this parent. #901 retains its isolated ADR 0134 repair on its own Draft head and that child evidence does not transfer to this parent. #934 and #936 remain Draft. #937 retains its post-write reauthorization rollback repair on its own Draft head. Keep descendants stacked until normal protected integration.
Do not self-approve, bypass required governance, force-push, weaken gates, transfer predecessor/child evidence, or use no-op trigger churn.