cub-scout Live Witness Watchlist

A repository document, rendered for the site. View source markdown.

Generated at: 2026-07-30T12:38:02.000Z UTC · source: committed helm-expt evidence for this rendered repository document.

This watchlist records cases where ordinary local live checks pass, but the stricter cub-scout witness finds a render/live mismatch that needs a capability, target, or lifecycle decision.

The first rows are strict witness findings on kind Kubernetes 1.30:

A follow-up capability witness proves the same rendered cert-manager and External Secrets CRDs preserve selectableFields on kind-kubernetes-1.35: selectableFields witness. The Kubernetes 1.30 rows remain on this watchlist because support claims stay profile-bound.

These are not Helm-vs-installer semantic parity defects. They are rendered/live target findings: the rendered objects include authored fields that the live API does not preserve with the same shape on this target profile. The route is to test the intended production Kubernetes version and feature-gate profile, model the server normalization explicitly, or create a profile-specific recipe or variant that does not claim those fields.

Machine-readable rows are in cub-scout-watchlist.csv.

Accepted normalization rules are recorded in normalization-rules.md. A strict witness block should appear either in this watchlist or in that normalization log.