Skip to content

fix: discover stage-level summary questions when units are skipped - #1110

Merged
apackeer merged 2 commits into
mainfrom
fix/stage-level-summary-questions
Sep 16, 2026
Merged

apackeer merged 2 commits into
mainfrom
fix/stage-level-summary-questions

Conversation

@apackeer

@apackeer apackeer commented Sep 10, 2026

Copy link
Copy Markdown
Contributor

Summary

Fix false SUMMARY_QUESTIONS_MISSING refusals when a workflow skips Units Generation and stores construction questions at stage level, as the infra scope does.

Summary-question discovery now follows the effective plan's artifact placement. Workflows that execute Units Generation continue to use each Unit's own questions and confirmation. Isolated stage reviews retain their own question placement even when the parent workflow skips Units Generation, while preserving the parent's confirmation and Change Control policy.

Regression coverage checks scope defaults, plan overrides, obsolete Unit directories, isolated review and completion, and missing, stale, rejected, or unauthorized confirmation evidence.

User experience

A user who answers construction questions, confirms the summary, and generates artifacts in the engine-selected directory can continue through the remaining stage checks. An isolated per-unit stage can also be reviewed without inheriting the parent workflow's artifact directory. Invalid confirmation evidence still blocks progress.

This restores the existing placement contract. It adds no commands, flags, or receipt-identity inputs and requires no configuration migration.

Validation

  • Reproduced the isolated review regression before the follow-up fix.
  • After rebasing onto the latest main, 129 cases across plan approval, summary authorization, ceremony policy, and isolated-stage suites passed with the relevant guards enabled.
  • Final generated distributions, deterministic packaging, all three TypeScript configurations, lint, and whitespace checks passed.
  • Full smoke/unit, integration, and deterministic/live SDK end-to-end checks have passing results across their runs and isolated reruns. Local fixture corrections were kept on a separate validation branch.
  • All four native Windows Kiro TUI cases passed with zero skips. Native Windows Claude startup and context-isolation checks also passed.
  • Broad live runs preceded the final upstream numeric gate-reply fix; the related suites and all contract checks were rerun after rebasing.
  • The broader live gate still exposes Codex configuration/policy, Kiro ACP workspace-forwarding, and Kiro IDE session-tracking failures outside this functional patch. Those failures are recorded separately; this description does not claim an entirely passing live matrix.

Checklist

  • I have reviewed the contributing guidelines.
  • I have performed a self-review of this change.
  • Changes have been tested.
  • The behavior and validation are documented in this description.

Acknowledgment

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of the project license.

@apackeer
apackeer force-pushed the fix/stage-level-summary-questions branch from d7d75bf to b020352 Compare September 16, 2026 04:49
@spraja08 spraja08 self-assigned this Sep 16, 2026
@apackeer
apackeer added this pull request to the merge queue Sep 16, 2026
Merged via the queue into main with commit 7998589 Sep 16, 2026
27 checks passed
@apackeer
apackeer deleted the fix/stage-level-summary-questions branch September 16, 2026 06:29
apackeer added a commit that referenced this pull request Sep 17, 2026
* origin/main:
  fix: make compiled dispatch exhaustive over TOOLS so native review briefs route (#1070) (#1115)
  test: skip POSIX gate-sensor fixture on Windows (#1208)
  fix: make /aidlc compose scopes durable across an engine reinstall (#1159)
  fix: refuse orphaned positionals at intent-create instead of storing a truncated description (#1114) (#1195)
  fix: route compiled gate-sensor dispatch through the engine namespace (#1166)
  fix: name the accepted values in review-path refusals (#1082) (#1194)
  fix: discover stage-level summary questions when units are skipped (#1110)
  fix: keep a JSON-scalar gate reply instead of parsing it away (#1186)
  fix: resolve the zero-Unit code-generation location in the traceability sensor (#1179)
  fix(config): one quiet line for --show, and honest source options on a copy-channel refresh (#1185)
  test(t238): normalize walkFiles separators in the invocation-surface selector (#1184)
  fix(config): Kiro provides its own model access, so the provider section has nothing to ask (#1183)
  chore: prepare 2.9.0 release (#1181)
  feat!: classic scope v1 parity with scope-owned ceremony switches (#1151)
  fix: publish changed previews without a daily cap (#1178)
  fix: separate Bun copy and native installer runtimes (#1174)
  fix: publish next-patch previews nightly (#1169)
  fix: keep review bookkeeping out of the artifact, and out of the reviewer's findings (#1160)
  fix: measure the review budget against the engine's own ordinal, not the caller's (#1158)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants