OPS.5:5 - Archetypal Grounding — Three Admission Replays

OPS.5:5.1 - PumpWorks controller service

PumpWorks-ControlServiceOps considers ReleaseCandidate-R42, FieldIncident-I73, SafetyQuestion-S19, ProviderChange-P8, and competing test-rig requests while maintaining weekly evidenced-release commitments. The current account exposes missing test T9, a safety-return gap, provider-access conflict, release authority, service consequences, and the current rig-access relation.

OPS.5 keeps the items separate. It can admit bounded incident-diagnosis Work for FieldIncident-I73 when identity, incident authority, evidence, performer, and access conditions obtain. It defers release continuation whose required T9 evidence or rig access is absent. It returns SafetyQuestion-S19 to the competent safety authority. A high rank, booked rig request, release candidate, or weekly horizon creates neither capacity, permission, Work, nor release. The first usable result is the item-by-item admission account and visible residual demand; the first honest blocker is the exact missing permission, evidence, access, or specialist return.

OPS.5:5.2 - Public-hospital emergency-care operation

One emergency-care operating function considers operating Work around exact patient and clinical-case subjects, changing observations, waiting, bed relations, service horizons, and current clinical and operating evidence. OPS.5 can return a bounded operating admission disposition only from authorized inputs. It does not turn a patient into a generic Work item or infer clinical priority from waiting age.

If consent, clinical evidence, competence, medical-safety, privacy, statutory permission, or another specialist condition is missing, the result names that blocker and its owner. Operations neither supplies treatment authority nor overrides clinical judgement. The first useful result is an authorized operating admission or an exact returned condition, not a dashboard state or automatic triage result.

OPS.5:5.3 - AI-assisted software operation

One software operation considers user issues, agent episodes, candidate changes, harness and tool access, traces, test results, human decisions, and release commitments. OPS.5 uses current context and tool capacity, bounded concurrency, integration burden, evidence needs, and established human or other authority to decide what may start.

A model score, available agent, free context window, generated patch, or scheduled episode does not by itself establish admission, permission, performed Work, or release. The result admits only the exact bounded Work whose identity, conditions, evidence, access, authority, and start limit obtain; it defers or returns the rest with retry conditions. Security, software assurance, provider capability, agenthood, and release authorization remain direct specialist returns.