FE-568: Requirements-review mode#36
Conversation
FE-568 Requirements-review mode
Synthesize the requirement set from the full knowledge layer, then let the user approve, edit, merge, reject, and add requirements through review turns. Acceptance
|
lunelson
changed the title
🤖 Augment PR SummarySummary: This PR grounds and hardens the first requirements-review loop, adding durable per-requirement review state and wiring requirements closure into the shared phase-close seam. Changes:
Technical Notes: Review state is derived from latest active-path review links (latest-action-wins), and requirements→criteria handoff reuses the existing 🤖 Was this summary useful? React with 👍 or 👎 |
![augmentcode[bot]](https://avatars.githubusercontent.com/in/1027498?s=60&v=4){kind=small}
| return outcome.closure_basis ?? null; | ||
| } | ||
|
|
||
| function hasRequirementsReviewCoverage(db: DB, projectId: number): boolean { |
There was a problem hiding this comment.
hasRequirementsReviewCoverage() is based on getRequirementEntitiesForProject(), which currently loads all requirement knowledge_item rows for the project; if a requirement was created on an inactive branch (no active-path review links), it will remain pending and can block requirements closeability/closure even when the active requirement set is fully reviewed.
Severity: medium
🤖 Was this useful? React with 👍 or 👎, or 🚀 if it prevented an incident/outage.
|
|
||
| const selectedReviewOptionPosition = | ||
| review.kind === 'requirement-approval' ? review.approveOptionPosition : review.rejectOptionPosition; | ||
| if (!selectedPositions.includes(selectedReviewOptionPosition)) { |
There was a problem hiding this comment.
recordRequirementReviewFromTurnResponse() records an approval/rejection as long as the review option position is included in selectedPositions, even if the response also selected other options (multi-select), which can persist explicit review state from an ambiguous reply and skew closeability/projection.
Severity: medium
🤖 Was this useful? React with 👍 or 👎, or 🚀 if it prevented an incident/outage.
lunelson
changed the base branch from
ln/fe-540-design-mode-commitment-exploration
to
graphite-base/36
…al writes, phase mismatch rejection on confirmations, correct meta label for closed phases Amp-Thread-ID: https://ampcode.com/threads/T-019d762a-383e-77bb-a53e-d9c8768dceaf Co-authored-by: Amp <amp@ampcode.com>
lunelson
force-pushed
the
ln/fe-568-requirements-review
branch
from
334d7bc to
9173768
Compare
1 participant
feat: ground the first requirements-review loop
feat: persist explicit requirement approval state
feat: close out minimal closure flow for reqs review