fix: make behaviour evaluation authoritative

Core fix: For cases with expectedBehaviours, reasoningQuality.status is now
set exclusively from behaviour evaluation results (required behaviour pass/fail).
Legacy concept checks remain visible as diagnostic-only metrics and do not
influence the authoritative result.

Key changes:
- Behaviour-based scoring determines reasoning status (passed/failed)
  instead of legacy concept literal matching
- Schema failure correctly forces not_evaluated (no vacuous truth)
- Saved live results re-evaluator preserves provenance metadata
- Classification tolerance map works bidirectionally for interchangeable types
- normalise() treats underscores as word characters, hyphens as spaces

Tests: 74 passing across both evaluator test suites
- tests/evaluator-behaviour-authoritative.test.mjs (47 tests, new)
- tests/evaluator-semantic.test.mjs (27 tests)

lib/reconstruction/prompt.js

@@ -45,7 +45,10 @@ Return ONLY the JSON object. No markdown, no explanation, no preamble.`;
 /** Load a versioned prompt from disk and substitute {{SCENARIO}} */
 async function buildV2Prompt(scenario) {
   try {
-    const content = await fs.readFile(join(PROMPTS_DIR, "reconstruct-v0.2.md"), "utf-8");
+    const content = await fs.readFile(
+      join(PROMPTS_DIR, "reconstruct-v0.2.md"),
+      "utf-8",
+    );
     return content.replace("{{SCENARIO}}", scenario);
   } catch {
     // Fall back to v0.1 prompt if v0.2 file is missing
@@ -69,6 +72,7 @@ export async function buildPrompt(scenario, version = "v0.2") {
       break;
   }
 
-  const strongJsonHint = "\n\nReturn ONLY a valid JSON object starting with { and ending with }. Do NOT include any text before the opening brace or after the closing brace. Do NOT wrap in markdown backticks.";
+  const strongJsonHint =
+    "\n\nReturn ONLY a valid JSON object starting with { and ending with }. Do NOT include any text before the opening brace or after the closing brace. Do NOT wrap in markdown backticks.";
   return { prompt: prompt + strongJsonHint, version };
 }

lib/reconstruction/schema.js

@@ -5,7 +5,12 @@ import { z } from "zod";
 // ──────────────────────────────────────────────
 
 export const confidenceEnum = z.enum(["low", "medium", "high"]);
-const importanceEnum = z.enum(["incidental", "supporting", "important", "critical"]);
+const importanceEnum = z.enum([
+  "incidental",
+  "supporting",
+  "important",
+  "critical",
+]);
 const expectedInfoValueEnum = z.enum(["low", "medium", "high"]);
 
 // ──────────────────────────────────────────────
@@ -24,8 +29,11 @@ export const reconstructionSchema = z.object({
   observations: z.array(itemSchemaV1),
   reportedClaims: z.array(
     itemSchemaV1.extend({
-      attributedTo: z.union([z.string().min(1), z.null()]).optional().nullable(),
-    })
+      attributedTo: z
+        .union([z.string().min(1), z.null()])
+        .optional()
+        .nullable(),
+    }),
   ),
   assumptions: z.array(itemSchemaV1),
   entities: z.array(itemSchemaV1),
@@ -35,7 +43,7 @@ export const reconstructionSchema = z.object({
       previousState: z.string().min(1),
       currentState: z.string().min(1),
       explanationStatus: z.string().min(1),
-    })
+    }),
   ),
   expectedButMissing: z.array(itemSchemaV1),
   presentButUnexpected: z.array(itemSchemaV1),
@@ -65,45 +73,53 @@ export const healthResponseSchema = z.object({
 // v0.2 — reasoning classification + reconstruction
 // ──────────────────────────────────────────────
 
-export const inputTypes = /** @type {z.ZodType<typeof import("@/lib/reconstruction/schema").INPUT_TYPE_VALUE>} */ (
-  z.enum([
-    "observed_problem",
-    "unexplained_change",
-    "contradiction",
-    "decision_request",
-    "causal_claim",
-    "reported_claim",
-    "fault_report",
-    "ambiguous_statement",
-    "question",
-    "desired_outcome",
-    "insufficient_context",
-    "other",
-  ])
-);
+export const inputTypes =
+  /** @type {z.ZodType<typeof import("@/lib/reconstruction/schema").INPUT_TYPE_VALUE>} */ (
+    z.enum([
+      "observed_problem",
+      "unexplained_change",
+      "contradiction",
+      "decision_request",
+      "causal_claim",
+      "reported_claim",
+      "fault_report",
+      "ambiguous_statement",
+      "question",
+      "desired_outcome",
+      "insufficient_context",
+      "other",
+    ])
+  );
 
-export const reasoningModes = /** @type {z.ZodType<typeof import("@/lib/reconstruction/schema").REASONING_MODE_VALUE>} */ (
-  z.enum([
-    "establish_baseline",
-    "identify_difference",
-    "reconstruct_transition",
-    "decompose_aggregate",
-    "validate_measurement",
-    "validate_claim",
-    "investigate_contradiction",
-    "clarify_meaning",
-    "decision_support",
-    "fault_investigation",
-    "identify_missing_information",
-    "test_possible_explanations",
-    "other",
-  ])
-);
+export const reasoningModes =
+  /** @type {z.ZodType<typeof import("@/lib/reconstruction/schema").REASONING_MODE_VALUE>} */ (
+    z.enum([
+      "establish_baseline",
+      "identify_difference",
+      "reconstruct_transition",
+      "decompose_aggregate",
+      "validate_measurement",
+      "validate_claim",
+      "investigate_contradiction",
+      "clarify_meaning",
+      "decision_support",
+      "fault_investigation",
+      "identify_missing_information",
+      "test_possible_explanations",
+      "other",
+    ])
+  );
 
 const evidenceRecordSchema = z.object({
   id: z.string().min(1),
   description: z.string().min(1),
-  evidenceType: z.enum(["direct_observation", "reported_statement", "interpretation", "assumption", "inferred_relationship"]),
+  evidenceType: z.enum([
+    "direct_observation",
+    "reported_statement",
+    "interpretation",
+    "assumption",
+    "inferred_relationship",
+  ]),
   source: z.string().optional(),
   attribution: z.string().nullable().optional(),
   confidence: confidenceEnum,
@@ -123,14 +139,14 @@ const reconstructionSchemaV2 = z.object({
       previousState: z.string().min(1),
       currentState: z.string().min(1),
       explanationStatus: z.string().min(1),
-    })
+    }),
   ),
   unexplainedTransitions: z.array(
     itemSchemaV1.extend({
       entity: z.string().min(1).optional(),
       previousState: z.string().min(1).optional(),
       currentState: z.string().min(1).optional(),
-    })
+    }),
   ),
   contradictions: z.array(itemSchemaV1),
   importantUnknowns: z.array(itemSchemaV1),
@@ -141,7 +157,7 @@ const reconstructionSchemaV2 = z.object({
       supportingEvidenceIds: z.array(z.string()),
       assumptionsRequired: z.array(z.string()).optional().default([]),
       confidence: confidenceEnum,
-    })
+    }),
   ),
 });