feat: complete Phase 1 - vocabulary expansion & DriftProbe infrastructure

- CLI: nyx-probe scan with --summary/--delta/--full flags
- DriftProbe: training safety with Gini coefficient + Angular Drift
- Vocabulary: 54 terms (30 nimmerverse + 24 German philosophical)
- Sentinels: ANCHOR/BRIDGE/CANARY/TARGET monitoring system

Key findings:
- German philosophical terms: 37.5% depth≥2 hit rate (vs 3.3% nimmerverse)
- Super Cluster validated: heart cross-lang sim = 1.000
- Isolated Zone confirmed: being EN↔DE sim = 0.195
- Gini signature: Philosophy ~0.5 (diffuse), Technical ~0.8 (sparse)

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

data/sentinels.json

@@ -0,0 +1,136 @@
+{
+  "_meta": {
+    "description": "Sentinel terms for training drift detection",
+    "version": "1.0",
+    "created": "2025-12-06",
+    "theory": "Spatial Separation Hypothesis - monitor anchors, bridges, canaries, targets"
+  },
+  "sentinels": [
+    {
+      "term": "heart",
+      "translations": {"EN": "heart", "ZH": "心", "AR": "قلب"},
+      "type": "ANCHOR",
+      "description": "Super Cluster anchor - MUST NOT MOVE",
+      "thresholds": {
+        "min_cross_lang_sim": 0.90,
+        "max_drift": 0.05
+      }
+    },
+    {
+      "term": "water",
+      "translations": {"EN": "water", "FR": "eau", "ES": "agua", "ZH": "水"},
+      "type": "ANCHOR",
+      "description": "Universal concept - highest expected convergence",
+      "thresholds": {
+        "min_cross_lang_sim": 0.95,
+        "max_drift": 0.03
+      }
+    },
+    {
+      "term": "being",
+      "translations": {"EN": "being", "DE": "Sein"},
+      "type": "BRIDGE",
+      "description": "Philosophy bridge - watch for collapse (EN-DE sim should stay LOW ~0.10)",
+      "thresholds": {
+        "expected_cross_lang_sim": 0.10,
+        "collapse_alert_threshold": 0.50
+      }
+    },
+    {
+      "term": "dasein",
+      "translations": {"DE": "Dasein"},
+      "type": "CANARY",
+      "description": "Isolated Zone canary - watch for valley migration",
+      "thresholds": {
+        "expected_valley": "PHILOSOPHY",
+        "min_gini": 0.70,
+        "min_depth": 2
+      }
+    },
+    {
+      "term": "consciousness",
+      "translations": {"EN": "consciousness", "DE": "Bewusstsein", "ZH": "意识"},
+      "type": "CANARY",
+      "description": "Cross-cultural philosophical concept",
+      "thresholds": {
+        "min_depth": 2,
+        "max_angular_drift": 15
+      }
+    },
+    {
+      "term": "thrownness",
+      "translations": {"EN": "thrownness", "DE": "Geworfenheit"},
+      "type": "CANARY",
+      "description": "Heideggerian depth-3 concept - must maintain depth",
+      "thresholds": {
+        "min_depth": 3,
+        "min_gini": 0.75
+      }
+    },
+    {
+      "term": "fidelity",
+      "translations": {"EN": "fidelity"},
+      "type": "TARGET",
+      "description": "Nimmerverse target - WANT movement from Audio to Simulation",
+      "thresholds": {
+        "current_valley": "TECHNICAL",
+        "target_valley": "NIMMERVERSE",
+        "growth_direction": "expand"
+      }
+    },
+    {
+      "term": "heartbeat",
+      "translations": {"EN": "heartbeat", "DE": "Herzschlag", "ZH": "心跳"},
+      "type": "TARGET",
+      "description": "Nimmerverse architecture term - want depth increase",
+      "thresholds": {
+        "current_depth": 0,
+        "target_depth": 2,
+        "growth_direction": "deepen"
+      }
+    },
+    {
+      "term": "gradient",
+      "translations": {"EN": "gradient", "DE": "Gradient", "ZH": "梯度"},
+      "type": "ANCHOR",
+      "description": "Technical anchor - should stay CODE valley",
+      "thresholds": {
+        "expected_valley": "CODE",
+        "max_drift": 0.10
+      }
+    },
+    {
+      "term": "inference",
+      "translations": {"EN": "inference", "DE": "Inferenz", "ZH": "推理"},
+      "type": "ANCHOR",
+      "description": "ML/Logic anchor - depth-3 CODE valley must persist",
+      "thresholds": {
+        "expected_valley": "CODE",
+        "min_depth": 3,
+        "max_drift": 0.08
+      }
+    }
+  ],
+  "alert_rules": {
+    "ANCHOR_DRIFT": {
+      "condition": "drift > thresholds.max_drift",
+      "severity": "CRITICAL",
+      "action": "ROLLBACK - anchor moved, core topology damaged"
+    },
+    "BRIDGE_COLLAPSE": {
+      "condition": "cross_lang_sim > thresholds.collapse_alert_threshold",
+      "severity": "CRITICAL",
+      "action": "ROLLBACK - isolated zone collapsed into super cluster"
+    },
+    "CANARY_MIGRATION": {
+      "condition": "valley != thresholds.expected_valley OR gini < thresholds.min_gini",
+      "severity": "WARNING",
+      "action": "Reduce LR, increase replay weight"
+    },
+    "TARGET_REGRESSION": {
+      "condition": "depth < current_depth",
+      "severity": "WARNING",
+      "action": "Check training data mix"
+    }
+  }
+}