🤖 Generated by the Daily AI Engineer
Evidence
world-at-ruin is the only one of 21 repositories whose IssueLabels CR is still
managementPolicies: [Observe]. Every other repo carries the full
[Observe, Create, Update, LateInitialize] set, so only world-at-ruin never has the canonical
taxonomy applied.
Measured on the live cluster (admin@prod, read-only):
issuelabels/world-at-ruin — Synced=True, Ready=True, spec.managementPolicies: ["Observe"]
- observed labels: 5 —
blocked, enhancement, next, released, roadmap
issuelabels/ksail (full-policy peer): 35 observed labels
- the shared patch in
deploy/labels/kustomization.yaml declares a 22-label canonical taxonomy
and targets every IssueLabels, so world-at-ruin already declares all 22 — it just never
applies them
Problem
World at Ruin is a first-class product, but its issues cannot be triaged honestly: it has no bug,
documentation, refactor, performance, security, needs triage or duplicate label. #124
("cave.tscn is unreferenced — delete it or mark it a preview harness") is a cleanup task with no
label that fits, so it sits untriaged — not an oversight, but the only available outcome.
The adoption gate is met
The CR's own comment states the promotion condition:
Promote to the full policy set (all except Delete) once Synced/Ready shows the observed state
matches the declared one.
Synced=True / Ready=True on the live resource satisfies exactly that. The repo was bootstrapped
2026-07-16 with the deliberate Observe-first stance (per the #43–#46 convention); the promotion step
was simply never taken afterwards.
Proposed change
Promote deploy/labels/world-at-ruin.yaml to [Observe, Create, Update, LateInitialize], matching
all 20 peers. Purely additive: all five existing labels (blocked, enhancement, next,
released, roadmap) are already members of the canonical 22, and no Delete policy is claimed,
so nothing is removed or recoloured.
Acceptance criteria
Size: XS. Part of #56.
Evidence
world-at-ruinis the only one of 21 repositories whoseIssueLabelsCR is stillmanagementPolicies: [Observe]. Every other repo carries the full[Observe, Create, Update, LateInitialize]set, so only world-at-ruin never has the canonicaltaxonomy applied.
Measured on the live cluster (
admin@prod, read-only):issuelabels/world-at-ruin—Synced=True,Ready=True,spec.managementPolicies: ["Observe"]blocked,enhancement,next,released,roadmapissuelabels/ksail(full-policy peer): 35 observed labelsdeploy/labels/kustomization.yamldeclares a 22-label canonical taxonomyand targets every
IssueLabels, so world-at-ruin already declares all 22 — it just neverapplies them
Problem
World at Ruin is a first-class product, but its issues cannot be triaged honestly: it has no
bug,documentation,refactor,performance,security,needs triageorduplicatelabel. #124("
cave.tscnis unreferenced — delete it or mark it a preview harness") is a cleanup task with nolabel that fits, so it sits untriaged — not an oversight, but the only available outcome.
The adoption gate is met
The CR's own comment states the promotion condition:
Synced=True/Ready=Trueon the live resource satisfies exactly that. The repo was bootstrapped2026-07-16 with the deliberate Observe-first stance (per the #43–#46 convention); the promotion step
was simply never taken afterwards.
Proposed change
Promote
deploy/labels/world-at-ruin.yamlto[Observe, Create, Update, LateInitialize], matchingall 20 peers. Purely additive: all five existing labels (
blocked,enhancement,next,released,roadmap) are already members of the canonical 22, and noDeletepolicy is claimed,so nothing is removed or recoloured.
Acceptance criteria
deploy/labels/world-at-ruin.yamluses the full policy set; the Observe-only caveat commentis replaced with the adoption evidence.
kubectl kustomize deploy/still builds and the repo's tests pass.Size: XS.
Part of #56.