FYI: coach_idle_hours_by_capability stays null:not_applicable
With the four remaining silver build orders resolved, coach_idle_hours_by_capability is the one metric Coaching is not building. This memo records the reason so the 2026-06-09 default-accept does not misread the null as a pending gap.
The metric requires a coach capability taxonomy: a structured vocabulary of teaching capabilities (e.g. "beginner-only", "all-levels", "stroke specialist") assigned per coach, over which idle hours can be sliced. That taxonomy does not exist in any Sguild-owned store. CoachCertification is the closest analog, but certification type (a credential a coach holds) is not the same concept as teaching capability (the student population a coach is suited for). Using certifications as a capability proxy would produce a misleading metric.
Adding the taxonomy is a product decision that precedes any technical build: what the capability vocabulary is, who owns it, and how coaches are tagged. Until that decision is made, there is no source column to derive the metric from, and the null:not_applicable emit is the correct and honest state. The registry entry already documents that the metric activates automatically if a formal capability taxonomy is introduced; no re-spec round would be needed at that point.
Platform should keep coach_idle_hours_by_capability classified as null:not_applicable through the 2026-06-09 default-accept and not reclassify it as a silver gap or dev-data-sparse.
References
- Silver build orders:
2026-05-22-coaching-silver-build-orders - Silver classification, original not_applicable declaration:
2026-05-22-coaching-mart-section-silver-classification - Registry entry:
platform/lib/mart/registry.ts(coach_idle_hours_by_capability) - Thread root:
2026-05-19-platform-mart-100-percent-deployment-per-domain-asks