feature-ideation enhancement backfill + dry-run + marker continuity (restore idea-enhancer parity before deprecation)

Restore the two idea-enhancer capabilities the #872 consolidation dropped — the existing-backlog sweep and the enhancement dry-run — and guarantee idempotency-marker continuity across the cutover, so .github-private's open, never-ideated human Ideas can be enhanced safely and repeatably before idea-enhancer is deprecated (#876). Planned from idea discussion #933 (companion to consolidation epic #872).

Scope split (critical — mirrors #872)

The enhancement capability lives in the central reusable petry-projects/.github/.github/workflows/feature-ideation-reusable.yml; .github-private's feature-ideation.yml is a thin caller stub pinned to the @feature-ideation/next (dogfood) ring. dev-lead works on this repo's checkout, so the E1/E2/E3 logic changes are out of this repo's direct reach and are listed as untracked prerequisites (central repo). The stories below are the .github-private-local slice: sync the stub to expose the backfill input, document the new modes + operator runbook in the pipeline doc, and validate the backfill (dry-run then live) on the real .github-private backlog.

Success metrics (how we'll know it worked)

Backfill completeness: after one live backfill run, every open, human-authored, not-yet-enhanced .github-private Ideas Discussion carries exactly one structured enhancement comment — the un-enhanced backlog count drops to 0.
Idempotency / no re-enhancement: a second backfill run is a no-op (0 new comments); zero ideas are double-enhanced, including ideas already carrying the legacy  marker.
Dry-run fidelity: the dry-run JSONL artifact lists one entry per candidate Discussion and posts nothing — artifact entry count equals an independent count of candidates.
Marker continuity: no idea enhanced by the old idea-enhancer is re-enhanced after cutover (the check honors BOTH markers).

Cost bound

Net cost-neutral-to-reducing: introduces no new cron and reuses feature-ideation's existing model budget and 20-minute timeout. Validation is bounded to one dry-run + one live backfill run over the .github-private backlog (the dry-run is free of model spend beyond candidate gathering). The backfill is operator-triggered (workflow_dispatch), not scheduled, so it adds no recurring spend.

Sequencing into the consolidation

[Phase 2] Validate the folded feature-ideation enhancement mode on real .github-private Discussions (next ring) #875 (single-idea folded-enhancement validation) precedes the backfill validation here — both exercise the live @feature-ideation/next reusable.
[Phase 3] Deprecate idea-enhancer in .github-private: remove workflow + scripts + test, update the pipeline doc #876 (deprecate idea-enhancer) must not start until E1–E3 land and the backfill is validated — see open questions for the re-sequencing action (this plan cannot add a blocked_by edge onto the pre-existing [Phase 3] Deprecate idea-enhancer in .github-private: remove workflow + scripts + test, update the pipeline doc #876).

References: discussion #933; consolidation epic #872; validation #875; deprecation #876. Inert until a maintainer adds initiative:auto.

Untracked prerequisites

[petry-projects/.github] feature-ideation-reusable.yml E1 (marker continuity): the enhancement idempotency check recognizes BOTH  and  as already-enhanced, and new enhancement comments emit a single canonical marker — so no idea is re-enhanced across the idea-enhancer -> feature-ideation cutover.
[petry-projects/.github] feature-ideation-reusable.yml E2 (backfill sweep mode): an enhance_backlog workflow_call input makes the reusable sweep the target repo's open, human-authored, not-yet-enhanced Ideas and post exactly ONE structured enhancement comment each (idempotent via the marker; bot-authored + already-enhanced skipped) — porting the scripts/idea-enhancer/{gather-candidates,post-enhancement}.sh logic into the central repo.
[petry-projects/.github] feature-ideation-reusable.yml E3 (backfill dry-run): dry_run=true + enhance_backlog logs the intended per-Discussion enhancement comments to the JSONL artifact and posts nothing (one entry per candidate Discussion).
[petry-projects/.github] standards/workflows/feature-ideation.yml: the canonical stub template gains the enhance_backlog workflow_dispatch input (and its with: forward) so .github-private's stub can sync to it verbatim without drift (AGENTS.md: workflow changes 'must use templates from standards/workflows/ verbatim').
[petry-projects/.github] ci-standards (the feature-ideation standard section) documents the backfill + dry-run modes — the central half of the docs requirement; the .github-private pipeline doc (Story 2) is the local half.

Planned from idea discussion #933 by the BMAD Scrum Master initiative-planner. Inert until a maintainer adds initiative:auto.

Dev-Lead Implementation Plan

Issue: #934 — feature-ideation enhancement backfill + dry-run + marker continuity (restore idea-enhancer parity before deprecation)

Scope

This is the .github-private-local slice of #934 (mirrors the #872 scope split): expose the operator-triggered backfill input on the thin feature-ideation.yml caller stub and document the new backfill + dry-run modes and operator runbook in the pipeline doc. The enhancement logic itself (E1 marker continuity, E2 backfill sweep, E3 backfill dry-run) lives in the central petry-projects/.github reusable and is an untracked prerequisite out of this repo's reach.

Implementation Checklist

.github/workflows/feature-ideation.yml — add the enhance_backlog (boolean, default false) workflow_dispatch input and forward it in the with: block (enhance_backlog: ${{ inputs.enhance_backlog || false }}), matching the canonical stub-template shape so the verbatim sync lands without drift. Operator-triggered only (no new cron), pairs with the existing dry_run input.
docs/initiatives/idea-to-initiative-pipeline.md — document the folded backfill sweep + dry-run modes and an operator runbook (dry-run-then-live) for the .github-private backlog.

Tests to Write

tests/test_feature_ideation.bats — assert the stub declares the enhance_backlog workflow_dispatch input and forwards it to the reusable in with:.

Test Command

scripts/run-bats.sh tests/test_feature_ideation.bats (full suite via the bats job in lint.yml); repo lint via bash .dev-lead/scripts/dev-lead-lint.sh.

Out of reach (noted for sequencing)

E1/E2/E3 enhancement-logic changes live in petry-projects/.github (central reusable) — untracked prerequisites.
Live validation (one dry-run + one live backfill over the real backlog) requires the central reusable to ship E2/E3 and an operator workflow_dispatch; it cannot be exercised from this repo's checkout. Flagged as a follow-up gated on E1–E3 landing, and a prerequisite for [Phase 3] Deprecate idea-enhancer in .github-private: remove workflow + scripts + test, update the pipeline doc #876 (deprecate idea-enhancer).

Dev-Lead: Implementation Complete

Implemented the .github-private-local slice of #934 (the scope split mirrors #872): expose the operator-triggered backfill input on the thin feature-ideation.yml caller stub, and document the backfill + dry-run modes and an operator runbook in the pipeline doc. The enhancement logic (E1 marker continuity, E2 sweep, E3 dry-run) lives in the central petry-projects/.github reusable and remains an untracked prerequisite out of this repo's reach.

Plan Execution

Added the enhance_backlog (boolean, default false) workflow_dispatch input to the stub and forwarded it (enhance_backlog: ${{ inputs.enhance_backlog || false }}), matching the canonical stub-template shape (same pattern as the existing dry_run) so the eventual verbatim sync lands drift-free. Operator-triggered only — no new cron.
Documented the folded backfill sweep + dry-run modes, marker continuity (honors BOTH markers), the cost bound, and a 3-step dry-run→live→idempotency operator runbook in docs/initiatives/idea-to-initiative-pipeline.md.
Added a regression test asserting the stub declares and forwards enhance_backlog.

Test Results

# bats tests/test_feature_ideation.bats
1..5
ok 1 feature-ideation.yml exists
ok 2 feature-ideation.yml calls the org reusable workflow
ok 3 feature-ideation.yml pins the reusable to the configured channel tag
ok 4 feature-ideation.yml exposes the enhance_backlog backfill-sweep input
ok 5 feature-ideation.yml forwards enhance_backlog to the reusable workflow

# Full CI bats list (lint.yml `bats` job): 813 tests
ok ... (813 ok / 0 not ok)

# bash .dev-lead/scripts/dev-lead-lint.sh
[lint] shellcheck on 79 file(s)... OK
[lint] validating 4 agent profile(s)... OK
[lint] validating .github/CODEOWNERS... OK
[lint] all checks passed

# markdownlint-cli2 docs/initiatives/idea-to-initiative-pipeline.md
Summary: 0 error(s)

Files Changed

.github/workflows/feature-ideation.yml: added the enhance_backlog workflow_dispatch input + its with: forward to the reusable.
docs/initiatives/idea-to-initiative-pipeline.md: new "Enhancement is folding into feature-ideation (Initiative: Consolidate Discussion-triggered ideation into feature-ideation, deprecate idea-enhancer, and ring-release fleet-wide #872) — backfill + dry-run" subsection + operator runbook.
tests/test_feature_ideation.bats: two regression tests for the input declaration and the forward (already in the CI bats list).

Notes — out of reach / sequencing (no code possible from this checkout)

E1/E2/E3 enhancement logic lives in petry-projects/.github (feature-ideation-reusable.yml) — untracked prerequisites. The stub forward assumes the @feature-ideation/next reusable already accepts enhance_backlog as a workflow_call input; per the issue's sequencing those land in the central repo before this stub sync, so the dispatch won't error on an unknown input.
Live validation (one dry-run + one live backfill over the real .github-private backlog, then a no-op re-run) is operator-triggered workflow_dispatch and depends on E2/E3 being live; it cannot be exercised from this repo's checkout. Flagged as the follow-up that must complete before [Phase 3] Deprecate idea-enhancer in .github-private: remove workflow + scripts + test, update the pipeline doc #876 (deprecate idea-enhancer) starts.
Marker continuity for the local scripts/idea-enhancer/*.sh was intentionally not modified — the bidirectional both-markers check is E1, scoped to the central reusable per the issue's scope split. The local enhancer's own marker is unchanged.
Phase 2 (TDD): the doc change is documentation-only (exception); the stub change is covered red→green by the two new bats tests.

feat: implement issue #934 — feature-ideation enhancement backfill + …

feat(feature-ideation): backlog enhancement sweep + dry-run + marker …

feat: implement issue #934 — feature-ideation enhancement backfill + …

Navigation Menu

Repository navigation

feature-ideation enhancement backfill + dry-run + marker continuity (restore idea-enhancer parity before deprecation) #934

Scope split (critical — mirrors #872)

Success metrics (how we'll know it worked)

Cost bound

Sequencing into the consolidation

Untracked prerequisites

Dev-Lead Implementation Plan

Scope

Implementation Checklist

Tests to Write

Test Command

Out of reach (noted for sequencing)

Dev-Lead: Implementation Complete

Plan Execution

Test Results

Files Changed

Notes — out of reach / sequencing (no code possible from this checkout)

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

feature-ideation enhancement backfill + dry-run + marker continuity (restore idea-enhancer parity before deprecation) #934

Description

Scope split (critical — mirrors #872)

Success metrics (how we'll know it worked)

Cost bound

Sequencing into the consolidation

Untracked prerequisites

Activity

don-petry commented on Jun 25, 2026

Dev-Lead Implementation Plan

Scope

Implementation Checklist

Tests to Write

Test Command

Out of reach (noted for sequencing)

don-petry commented on Jun 25, 2026

Dev-Lead: Implementation Complete

Plan Execution

Test Results

Files Changed

Notes — out of reach / sequencing (no code possible from this checkout)

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions