What Runtime Does
Runtime applies the governed asset at the moment work needs guidance. It receives a bounded request, resolves the active publication for the defined activation target, evaluates the relevant compiled intelligence, and produces an answer or guidance state.
This makes Runtime different from a chatbot session. The center of the interaction is not whatever a model can infer from temporary context. The center is what the organization has governed, compiled, activated, and authorized for that context.
Runtime Fails Closed
SALARA's deterministic Runtime records statuses such as ANSWER_ESTABLISHED, ANSWER_NOT_ESTABLISHED, NO_ACTIVE_ORGANIZATIONAL_INTELLIGENCE, RUNTIME_SCOPE_MISMATCH, INVALID_ACTIVE_SELECTION, and RUNTIME_FAILED_CLOSED.
Those states are important because enterprise systems need visible limits. When the active governed asset does not establish an answer, Runtime should not quietly invent authority.
Why Runtime Matters
Runtime is where Organizational Intelligence becomes useful without losing accountability. A person or system can receive guidance, and the result can still be traced back to the publication, activation, compiled artifact, Assertions, evidence, and ledger records behind it.
That traceability is what lets Proof Space demonstrate SALARA working: readers can ask, inspect, and replay instead of only reading claims about the architecture.