Background
PR #120 stopped propagating Samples/HAR-Original/ and .dogfood-output/ to consumer .gitignore files (the former is now owned by the api-wrapper-scaffold skill; the latter is upstream-only). However, the .gitignore merge in Merge-FileFromUpstream is union-only -- it never removes entries. Consumers that synced before PR #120 still have both lines in their .gitignore.
The entries are inert (the directories don't exist in typical consumers), but they are noise that a maintainer reviewing their .gitignore may stumble over.
Proposed change
Add a one-paragraph note to README.md under the bootstrap / sync section explaining:
Acceptance
- README section exists and clearly lists the two legacy entries with a one-line rationale each.
Out of scope
- Automating removal in
Pull-SDLC.ai.ps1 (would require a "deprecated entries" mechanism in the merge engine -- high complexity, low value).
Context
Spawned from PR #120 follow-up review.
Background
PR #120 stopped propagating
Samples/HAR-Original/and.dogfood-output/to consumer.gitignorefiles (the former is now owned by theapi-wrapper-scaffoldskill; the latter is upstream-only). However, the.gitignoremerge inMerge-FileFromUpstreamis union-only -- it never removes entries. Consumers that synced before PR #120 still have both lines in their.gitignore.The entries are inert (the directories don't exist in typical consumers), but they are noise that a maintainer reviewing their
.gitignoremay stumble over.Proposed change
Add a one-paragraph note to
README.mdunder the bootstrap / sync section explaining:Samples/HAR-Original/should be kept (or re-added by theapi-wrapper-scaffoldskill) only by consumers that actually use the API-wrapper scaffold..dogfood-output/is upstream-only and never needs to be in a consumer.gitignore.Acceptance
Out of scope
Pull-SDLC.ai.ps1(would require a "deprecated entries" mechanism in the merge engine -- high complexity, low value).Context
Spawned from PR #120 follow-up review.