Core Concepts and Threat Actors

Defensible only if the classes track a position your estate does not hand out. Re-cut the boundary from advisory wording to who already holds the position, add chain and population escalation clauses, and keep a window on every class.

Fund the identifier. It is usable on reading by every account holder, so a crew with no exploit-development budget can use it today, while converting the parser defect needs funded research they would have to buy. Severity ranks impact, not distance.

A Class label is a vocabulary term, not a work item. Three hundred rows sharing it span unrelated resources and unrelated fixes, so closing them together closes nothing. Re-map the dominant ones to Base instead.

It justifies funding what patience and return require: access paths that stay closed after eviction, and authentication a stolen secret cannot satisfy. It does not justify buying against exotic capability, because the label asserts sponsorship, not novel exploits.

Answer without arguing about honesty. Trust addresses at most one of three insider problems and none of the loss from carelessness or a taken session, so move the decision from who holds a role to what that role may reach, and price the smallest scope cut you can defend.

Fund the footprint first: it is the only one of the two that removes an order of targeting rather than redirecting it. Take supplier terms at contract renewal, and name the name-first order as accepted, not covered.