feat(harness): phase.py rewrite + effective_decisions (D15 core fix, F2)

The single most important architectural fix vs ChironeWp3: ALL helpers consult
effective_decisions() instead of raw list_decisions(), so the reopen-aware view is
consistent everywhere (fixes the bug where approved_ctes/advance_problems conflated
stale pre-reopen decisions with new ones).

Model (corrected during TDD):
- current_phase folds the audit (excluding retracted) with guard 'n == cur' --
  already reopen-aware (old phase_approved:N after reopen to M<N don't advance).
- effective_decisions = decisions whose phase <= current_phase. A sql_approved at
  phase 7 is stale when current_phase=4 after a rollback to F4, even if in the ledger.
  Rollback to F4 does NOT invalidate decisions of phases 1-3 (they stay effective).
- decision_retracted markers excluded (audit-only).

Also: session/models.py ported (SchemaLinking + Candidate with grounded_values D14a
+ concept_formulas D14b). MAX_PHASE/PHASE_NAMES read from workflow.yaml via
load_workflow() (no duplication). Strada 2: ladder if-phase-N kept for now,
generic prerequisites evaluator (F2 full) deferred.

7 phase tests + 20 total passing.

harness/nsp/session/models.py

@@ -0,0 +1,64 @@
+from datetime import datetime
+from typing import Literal
+
+from pydantic import BaseModel, Field, ConfigDict
+
+
+# Stub locale di _YamlModel (in ChironeWp3 vive in mschema/models.py).
+# Qui serve solo come base con populate_by_name; mschema/ sara' portato nel Task A9.
+class _YamlModel(BaseModel):
+    model_config = ConfigDict(populate_by_name=True)
+
+
+class SessionManifest(_YamlModel):
+    id: str
+    created_at: datetime
+    status: Literal["open", "closed", "finalized"] = "open"
+    question: str
+    database: str
+    db_schema: str = Field(alias="schema")
+    # D12/D15: autore della sessione (auth) e versione del workflow usato.
+    author: str | None = None
+    summary: str | None = None
+    updated_at: datetime | None = None
+    updated_by: str | None = None
+    schema_version: int | None = None
+
+
+class Candidate(BaseModel):
+    kind: Literal["table", "column"]
+    name: str
+    signals: dict = {}
+    evidence: list[str] = []
+    decision: Literal["promoted", "excluded", "pending"] = "pending"
+    decision_seq: int | None = None
+    # D14a: valori citati nella domanda ancorati a questa colonna/tabella.
+    grounded_values: list[dict] = []
+
+
+class Join(BaseModel):
+    from_: str = Field(alias="from")
+    to: str
+    source: str = ""
+    decision: Literal["promoted", "excluded", "pending"] = "promoted"
+    decision_seq: int | None = None
+
+    model_config = {"populate_by_name": True}
+
+
+class ExcludedItem(BaseModel):
+    kind: Literal["table", "column"]
+    name: str
+    decision_seq: int | None = None
+
+
+class SchemaLinking(BaseModel):
+    question: str
+    candidates: list[Candidate] = []
+    joins: list[Join] = []
+    excluded: list[ExcludedItem] = []
+    open_questions: list[str] = []
+    # D14b: formule di concetto approvate, parte dello schema-linking.
+    concept_formulas: list[dict] = []
+
+    model_config = {"extra": "forbid"}