A classification of the coordination-status of declared coordination-scheme records, not a verdict on which institutions are right. The corpus fixes what a record is of (Paper 13), what testimony transmits between embedded observers (Paper 14), what elevates a relation to law (Paper 17), when a claim is true (Paper 20), what an action is for a single embedded agent (Paper 23), the per-agent responsibility verdict (Paper 25), the group-level allocation of responsibility (Paper 26), the grounding of value (Paper 27), and when a postulate may be adopted (Paper 63) — but nowhere types the history-internal, system-relative coordination-status of a declared multi-agent coordination-scheme record. The sharpest display of the gap: two declared schemes over one declared agent roster and case family, with identical recorded conduct — of which one satisfies the declared rule-generality criterion while the other violates it (a general rule versus an extensionally identical ad-hoc case table; the difference criterion-indexed, never authority-issued). While no such layer exists, a Global Planner — an external design authority whose operative core is fixing which coordination scheme counts as coordinating — works by default; it is distinct from Paper 23's Global Controller (a single agent's practical omnipotence) and from Paper 26's Global Allocator (an authoritative apportioner of responsibility). Unlike the fresh openings of Papers 75, 78, and 79, this seat is the corpus's own named backlog row, honored: the backlog has carried "Coordination / institutions without a global planner — many-agent extension of Action (23) and Testimony (14)" by name, and the extension consists in consuming those parents' typed records as declared inputs, never in re-typing action or transmission. Five type facts stand in print and bind the paper: "Typed is not adopted; classified is not endorsed." / "A coordination-status is a type, not a world-side verdict on which institution is right." / "A criterion-satisfaction status is a checked certificate, not a decided coordination success." / "A status, not an action verdict; a status, not an allocation." / "Coordination-standing is relative to a declared system and a declared criterion, never simpliciter." A coordination-scheme record is not an institution's justification. It is a recorded pairing of a declared agent set, a declared scheme, and declared compliance records — never a presupposed institutional fact; and compliance talk refers to recorded response events and their re-computed compliance checks, not world-side successful coordination. The construction is a coordination-status grounding object C(r) over a form-fixed bearer r = (roster, case family, scheme, recorded-response store) — registered by an independent Paper-13 record Reg(r), which strict recovery returns and which is never a scheme-provenance entry — wrapping the declared systems, the non-empty coordination-criteria store with checker rules and a six-condition criterion-admissibility rule (anti-fiat and discrimination-witnessed with complete-assignment witnesses, verdict-free, checker-guarded, pre-declared before any certificate is evaluated, and practice-provenanced status-neutrally — recorded doings, never recorded standing verdicts), the two-part provenance store (scheme-provenance and criterion-provenance, separately typed), the use record with use-relative required-criteria sets, and a scheme-individuation grain. On it, the official typing output CoordinationTyping(r; S) = (CoordCritStatus, a routing summary, C, P) is assigned by a certificate-based decidable test over finite declared obligation sets: the per-criterion status table is itself the first output component — a deliberate design difference from Paper 79's displayed-diagnostic CritStatus — with each criterion decided in five ordered stages (conflict / violated / satisfied / partial / open: a checked violation of a named clause defeats the conjunctive criterion; satisfaction may be established by a full certificate or by checked partial covers jointly exhausting the obligation set; partiality is reserved for genuinely undecided remainders); the criteria store carries a coordination-facing assignment-predictability criterion alongside representational forms; the C-component is derived from a per-cell table with checked non-compliance a first-class route (typed-compliance-violated-in-S); the P-component is structured as a mode classification plus a transfer sub-component of checked transfer certificates; and "satisfied" in the cell names abbreviates "a declared certificate checks against the declared admissible criterion c," never "the scheme really coordinates." The contribution beyond "compliance-in-S restated" — the 23/26-adjacency collapse — is exhibited, not asserted (the K0 guard, two-pronged): the paper displays two fully declared schemes over one declared roster, one declared case family, and one identical declared recorded-response store, with the typing differing at full strength — satisfied against violated — the failed scheme failing a named clause of a declared admissible criterion by a checked violation-witness; plus witnesses that (C, P) is not a function of the criterion-satisfaction component, and a worked mixed instance (satisfied at the assignment-predictability criterion while violated at rule-generality) exercising the routing convention; and the components are status-like rather than metadata: a declared consumption/transfer condition discipline, grounded non-circularly in the recorded use-classes, gates reliance by use-relative required-criteria sets and transfer by the typed transfer sub-component (CheckedTransferRule with its declared domain), never by mode labels. The multi-axis typing locates the rational-selection reading (Lewis), the focal-point display (Schelling), the recorded-practice counterpressure (Ostrom), the norm-grammar display (Bicchieri), and the constructionist face (Searle) as located displays and pressure-tests — none of these sources asserts the external-Planner thesis, and none is an evaluation input. The ceiling is declared and patrolled: typing is not adopting; classifying is not endorsing. The substantive which-institution-is-right question is declined either way — neither optimal-design essentialism nor institutional deflation; action, individual verdicts, allocation, transmission, adoption, value, lawhood, and truth stay at their owners' seats; strict recovery of Paper 13's per-record semantics is exact (Lemma R-co), over the registration record and all four event stores. The landing is CoordinationTyping-layer = scoped, with a null via K0 an honest branch and a Planner-dependence theorem (K3) declared and patrolled. Nothing is adopted; no truth, satisfaction, or world-side institutional-rightness verdict is claimed; no action is typed; nothing is allocated.
Tomoyuki Uchida (Fri,) studied this question.