feat(callcenter): odoo savant families + Layer-2 axioms + style wiring (D-ODOO-SAV-1/2/3)

Picks up the cross-repo handover boundary declared in
.claude/odoo/SAVANTS.md §"lance-graph handover boundary": the woa-rs
session defined the 25-Savant roster + delegation tuples; this lands the
lance-graph side's first three deliverables. Reasoner impls (D-ODOO-SAV-4)
are scoped in the plan and deferred to a follow-up PR.

Integration plan:
- .claude/plans/odoo-savant-reasoners-v1.md (new) + INTEGRATION_PLANS.md
  prepend, per board-hygiene.

D-ODOO-SAV-1 — two new OGIT families:
- FAMILY_PRODUCT_CATALOG = 0x64 (100), FAMILY_HR_FOUNDATION = 0x90 (144)
  in odoo_alignment.rs + family_registry.ttl rows.
- CORRECTION of the SAVANTS.md proposal: it named 0x63 for ProductCatalog,
  but 0x63 (=99) is already ogit:MRORepair in the authoritative registry.
  lance-graph is authoritative for the registry, so ProductCatalog gets the
  next free commercial-cluster byte 0x64 (=100). Documented on the constant
  + in family_registry.ttl + guarded by a test.
- Seed rows: product.pricelist→schema:PriceSpecification,
  product.pricelist.item→schema:UnitPriceSpecification, uom.uom→qudt:Unit
  (ties into the bO-4 QUDT spine), hr.employee→vcard:Individual,
  hr.department→org:OrganizationalUnit, hr.job→org:Role,
  hr.contract→fibo:Contract (base only; payroll is Enterprise/absent).
- product.template / product.product STAY on BillingCore 0x61 (billable
  items, not catalogue structure) — no drift of the existing seed/tests.

D-ODOO-SAV-2 — Layer-2 alignment axioms (data/ontologies/odoo/alignment/
odoo-to-foundation.ttl, new):
- Honest semantic alignment for account.account.tag → skos:Concept.
- Documented-None-with-rationale for the genuinely cross-cutting classes
  (stock.move/rule/orderpoint, account.analytic.distribution.model): no
  foundry family minted (Option B — None stays None without an honest
  pivot), matching their family=None status in the savant roster.
- Public OWL vocabularies referenced, never modified.

D-ODOO-SAV-3 — StyleCluster wiring:
- family_default_style(OgitFamily) -> Option<StyleCluster> + resolve_odoo_style(class).
  Pins 0x60→Direct, 0x61/0x62/0x64→Analytical, 0x80/0x90→Empathic, 0x81→Direct
  per SAVANTS.md inherited-style map.
- dolce_odoo gains a `uom.` → Quality suffix rule.

Tests: 144 lance-graph-callcenter tests pass (19 odoo_alignment incl. 9 new:
new family bytes free in registry, catalogue→ProductCatalog, billable-items-
stay-BillingCore deviation guard, hr→HRFoundation, live-table end-to-end for
both basins, classes-still-None-after-D1, family_default_style pins,
resolve_odoo_style chains, uom dolce rule). Updated the pre-existing
unmapped_classes_return_none test (hr.employee no longer None). Clippy clean
(only pre-existing CausalEdge64 deprecations elsewhere).

Author: claude

.claude/board/INTEGRATION_PLANS.md

@@ -1,3 +1,25 @@
+## 2026-05-27 — odoo-savant-reasoners-v1 (lance-graph side of the Odoo richness harvest: 2 new OGIT families + Layer-2 axioms + StyleCluster wiring + 5 Reasoner impls)
+
+**Status:** PROPOSAL (picks up the cross-repo handover boundary in `.claude/odoo/SAVANTS.md` §"lance-graph handover boundary"). woa-rs defined the 25-Savant roster + delegation tuples; lance-graph implements (a) Reasoner impls, (b) 2 new families + Layer-2 alignment axioms for the `None` classes, (c) StyleCluster wiring.
+**Confidence:** HIGH on (b)/(c) — additive extensions of `odoo_alignment.rs` seed + alignment TTLs. MED on (a) — Reasoner dispatch shape (one impl per ReasoningKind vs savant-config registry) pinned but needs a review pass.
+**Plan file:** `.claude/plans/odoo-savant-reasoners-v1.md`
+**Predecessors:** PR #412 (odoo hydrator + dolce_odoo classifier + ODOO slot 50), PR #413 (briefing pack).
+**Anchored iron rules:** I-VSA-IDENTITIES (savant = Layer-2 role catalogue), AGI-as-glove, board-hygiene, Iron Rule 1 (no brain-crate in customer binary), Iron Rule 7 (verhaltens-bewahrend — reasoner output is suggestion-only).
+
+### Scope
+Group B — `0x63 ProductCatalog` (Analytical) + `0x90 HRFoundation` (Empathic) families + Layer-2 alignment axioms for `stock.*` / `account.analytic.distribution.model` / `account.account.tag` (land on existing pivot where honest, else documented `None`). Group C — `StyleCluster` per family (field-or-sidecar). Group A — `SavantConclusion` + 5 `Reasoner` impls (one per `ReasoningKind`) in lance-graph-callcenter, dispatching on evidence + family style, `InferenceType::default_strategy()` → QueryStrategy, NarsTruth evidence fusion.
+
+### Deliverables
+D-ODOO-SAV-1 two new families + seed rows + family_registry.ttl · D-ODOO-SAV-2 Layer-2 alignment axioms TTL · D-ODOO-SAV-3 StyleCluster per family · D-ODOO-SAV-4 5 Reasoner impls (gated on dispatch-shape review, own PR).
+
+### Execution
+D-ODOO-SAV-1/2/3 additive + low-risk → first PR (this session). D-ODOO-SAV-4 → follow-up PR after `/code-review` on dispatch shape. Plan + INTEGRATION_PLANS prepend land with D-ODOO-SAV-1.
+
+### Invariants
+Option B (inherit existing slots; new families are genuine basins not per-class mints; `None` stays `None` w/o honest pivot) · public OWL pristine (axioms are NEW TTL) · savant = Layer-2 catalogue · reasoner output = suggestion (guard stays in woa-rs) · impls in callcenter behind contract `Reasoner` trait.
+
+---
+
 ## 2026-05-27 — atom-mailbox-substrate-v1 (ladder-serves-mailbox: atoms→styles→personas, quorum projection, counterfactual mantissa, AriGraph hot/cold/tombstone)
 
 **Status:** PROPOSAL (implements `EPIPHANIES.md` E-LADDER-SERVES-MAILBOX; extends `rung-persona-orchestration-v1` D-PERSONA-1 downward into the atom layer and outward into the mailbox lifecycle).

.claude/plans/odoo-savant-reasoners-v1.md

@@ -0,0 +1,123 @@
+# odoo-savant-reasoners-v1 — lance-graph side of the Odoo richness harvest
+
+> **Status:** PROPOSAL. Picks up the explicit cross-repo handover boundary
+> declared in `lance-graph/.claude/odoo/SAVANTS.md` §"lance-graph handover
+> boundary": the woa-rs session defined the 25-Savant roster + delegation
+> tuples + call sites + evidence schemas (all contract-level, BBB-allowed);
+> the **lance-graph side** must now implement (a) the `Reasoner` impls, (b)
+> the two new OGIT families + Layer-2 alignment axioms for the `None`
+> classes, (c) the `StyleCluster` wiring.
+>
+> **Confidence:** HIGH on (b)/(c) — concrete additive extensions of the
+> existing `odoo_alignment.rs` seed table + alignment TTLs. MED on (a) — the
+> `Reasoner` dispatch shape is an architectural decision (one impl per
+> `ReasoningKind` vs a savant-config registry) that this plan pins but which
+> needs a review pass before the larger build.
+>
+> **Predecessors:** PR #412 (odoo hydrator + `dolce_odoo` classifier + ODOO
+> slot 50), PR #413 (briefing pack import). Reads on:
+> `lance-graph-contract::reasoning` (Reasoner / ReasoningKind / ReasoningContext),
+> `lance-graph-callcenter::{family_table, odoo_alignment, unified_bridge}`,
+> `lance-graph-contract::thinking::StyleCluster`.
+>
+> **Anchored iron rules:** I-VSA-IDENTITIES (savant = Layer-2 role catalogue,
+> content in tables not bundles), the AGI-as-glove doctrine (savant dispatch
+> rides `MetaColumn`/`EdgeColumn`, no new service), board-hygiene.
+
+## Scope
+
+Three deliverable groups, matching SAVANTS.md (a)/(b)/(c):
+
+### Group B — two new OGIT families + Layer-2 alignment axioms (lowest risk, first)
+
+The roster needs homes for classes that resolve to `None` today:
+
+- **`0x63 ProductCatalog`** (basin: product catalogue + pricelist + UoM;
+  default style **Analytical**). Lands `product.template`,
+  `product.pricelist`, `product.pricelist.item`, `uom.uom`,
+  `product.category`. Currently `product.*` inherits `0x61 BillingCore`
+  slot 1 via the prefix fallback — ProductCatalog promotes the
+  catalogue/pricing concepts to their own basin so L8's
+  `PricelistAssignmentAgent` has a real family instead of `None`.
+- **`0x90 HRFoundation`** (basin: employee / org; default style
+  **Empathic**). Lands `hr.employee`→`vcard:Individual`,
+  `hr.department`→`org:OrganizationalUnit`, `hr.job`, `hr.contract` (base
+  only — payroll engine is Enterprise/absent, flagged). All `hr.*`
+  resolve `None` today.
+- **Layer-2 alignment axioms** for the remaining `None` classes that do
+  NOT get a new family (they stay cross-cutting): `stock.*` (stock.move,
+  stock.rule, stock.warehouse.orderpoint), `account.analytic.distribution.model`,
+  `account.account.tag`. These get `owl:equivalentClass` / `rdfs:subClassOf`
+  rows in `data/ontologies/odoo/alignment/` pointing at existing pivots
+  (e.g. `stock.move`→`schema:MoveAction`-style, analytic→`fibo:CostCenter`)
+  so `resolve_odoo` can land them on an existing family rather than minting
+  one. Where no honest pivot exists, the class stays `None` and the plan
+  records WHY (genuinely cross-cutting, needs a runtime SPO-G edge not a
+  static family).
+
+### Group C — StyleCluster wiring
+
+`FamilyEntry` carries `dolce_marker` + `owl_characteristics` but NOT the
+default `StyleCluster`. SAVANTS.md inherits the style from the family. Add a
+`default_style: StyleCluster` field to `FamilyEntry` (or a sidecar
+`family_style(OgitFamily) -> StyleCluster` map if adding a field churns too
+many call sites — decide in review). Seed:
+0x60→Direct, 0x61→Analytical, 0x62→Analytical, 0x63→Analytical,
+0x80→Empathic, 0x81→Direct, 0x90→Empathic.
+
+### Group A — Reasoner impls (largest, scoped here, built in a follow-up PR)
+
+The 25 savants collapse onto **5 `ReasoningKind` discriminants** ×
+**5 `InferenceType` strategies**. Decision to pin: implement **one
+`Reasoner` per `ReasoningKind`** (CustomerCategory, PostingAnomaly,
+NextBestAction, InvoiceCompleteness, Other) that dispatches on
+`ReasoningContext.evidence` + the resolved family's style, rather than 25
+separate impls. Each `reason()` selects its `QueryStrategy` from
+`InferenceType::default_strategy()` (already mapped:
+Deduction→CamExact, Induction→CamWide, Abduction→DnTreeFull,
+Revision→BundleInto, Synthesis→BundleAcross) and fuses evidence via the
+savant's `SemiringChoice` (NarsTruth = the common case). Conclusion type:
+a truth-weighted `SavantConclusion { suggestion, confidence: NarsTruth,
+rationale }` that woa-rs applies as a **suggestion only**, never an
+un-guarded write (Iron Rule 7, verhaltens-bewahrend). Home crate:
+`lance-graph-callcenter` (BBB-allowed, already owns `odoo_alignment.rs`).
+
+## Deliverables
+
+- **D-ODOO-SAV-1** (Group B): `0x63 ProductCatalog` + `0x90 HRFoundation`
+  family constants + seed rows in `odoo_alignment.rs`; `data/family_registry.ttl`
+  family declarations; product.*/hr.* seed rows. Tests: resolve
+  product.template→0x63, hr.employee→0x90, end-to-end against live table.
+- **D-ODOO-SAV-2** (Group B): Layer-2 alignment axioms TTL for stock.*,
+  analytic.distribution.model, account.account.tag — land on an existing
+  pivot where honest, else documented `None`-with-rationale. Tests: the
+  newly-aligned classes resolve; the genuinely-cross-cutting ones still
+  return `None` with a recorded reason.
+- **D-ODOO-SAV-3** (Group C): `StyleCluster` per family. Field-or-sidecar
+  decision in review; seed the 7 families; test each family→cluster.
+- **D-ODOO-SAV-4** (Group A): `SavantConclusion` + 5 `Reasoner` impls in
+  `lance-graph-callcenter`, dispatching on `ReasoningKind` + evidence +
+  family style. Per-impl tests with synthetic `EvidenceRef` batches.
+  **Gated on a review pass of the dispatch shape.** Its own PR.
+
+## Execution
+
+D-ODOO-SAV-1/2/3 are additive + low-risk → ship together in the first PR
+(this session). D-ODOO-SAV-4 (Reasoner impls) is the architectural piece →
+scoped here, built in a follow-up PR after a `/code-review` pass on the
+dispatch shape. Board-hygiene: this plan + INTEGRATION_PLANS prepend land
+in the same commit as D-ODOO-SAV-1.
+
+## Invariants
+
+- Option B holds: odoo classes INHERIT existing slots via OWL pivot; new
+  families 0x63/0x90 are genuine new basins (product catalogue, HR), not
+  per-class mints. `None` stays `None` when no honest pivot exists.
+- Public OWL/RDF sources stay pristine — alignment axioms are NEW TTL in
+  `data/ontologies/odoo/alignment/`, not edits to upstream vocabs.
+- Savant = Layer-2 role catalogue (I-VSA-IDENTITIES): identity in the
+  family table, content (evidence/rules) in Arrow/SPO, never bundled.
+- Reasoner output is a suggestion; the deterministic guard stays in woa-rs
+  (verhaltens-bewahrend). lance-graph implements the ambiguous core only.
+- No brain-crate in the customer binary (Iron Rule 1): impls live in
+  `lance-graph-callcenter` behind the contract `Reasoner` trait.

crates/lance-graph-callcenter/data/family_registry.ttl

@@ -161,6 +161,16 @@ ogit:MRORepair
     ogit.meta:superDomain "WorkOrderBilling" ;
     ogit.meta:familyId    "99"^^xsd:unsignedByte .
 
+# ProductCatalog 0x64 — product catalogue + pricelist + UoM basin (odoo
+# richness harvest, plan odoo-savant-reasoners-v1 D-ODOO-SAV-1). NOTE: the
+# woa-rs SAVANTS.md proposal named 0x63 for this, but 0x63=99 is already
+# MRORepair above; the lance-graph authoritative registry assigns the next
+# free commercial-cluster byte 0x64=100 instead.
+ogit:ProductCatalog
+    a ogit:FamilyNamespace ;
+    ogit.meta:superDomain "WorkOrderBilling" ;
+    ogit.meta:familyId    "100"^^xsd:unsignedByte .
+
 # ── OSINT basins 0x70..=0x74 ─────────────────────────────────────────────────
 
 ogit:OsintMaltego
@@ -198,6 +208,16 @@ ogit:SmbFoundryTaxDeclaration
     ogit.meta:superDomain "SMB" ;
     ogit.meta:familyId    "130"^^xsd:unsignedByte .
 
+# ── HR basin 0x90 ────────────────────────────────────────────────────────────
+# HRFoundation 0x90 — employee / org / job / base-contract basin (odoo
+# richness harvest, plan odoo-savant-reasoners-v1 D-ODOO-SAV-1). hr.* resolve
+# None today; this basin gives them a home. Payroll ENGINE is odoo Enterprise
+# (absent) — only the hr base data/structure is aligned here.
+ogit:HRFoundation
+    a ogit:FamilyNamespace ;
+    ogit.meta:superDomain "HR" ;
+    ogit.meta:familyId    "144"^^xsd:unsignedByte .
+
 # ── SMB BSON-shape basins 0xA0..=0xAD ────────────────────────────────────────
 # OQ-4 resolution (locked 2026-05-13): ogit.SMB.bson: sub-namespace carries the
 # 14 BSON-shape entities.  Slot range 0xA0..=0xAD (160..=173) is unconflicted