feat(example): add per-entity LLM evaluations for 985 WoN entities (S3.3)

Batch evaluation of all 988 entities via OpenRouter. 984 succeeded on
first pass; 3 failed (network errors). eval-summary --update-metrics
written with per_entity_mean=3.9556.

Viability dashboard: 6/6 PASS
  redundancy_ratio   0.0061  (max 0.10)
  coverage_ratio     0.6190  (min 0.40)
  coherence_comps    0.0000  (max 3)
  consistency_cycles 0.0000  (max 0)
  granularity_entropy 2.6748 (min 1.0)
  per_entity_mean    3.9556  (min 3.5)

Dimension breakdown (mean across 985 entities):
  definition_precision  3.62
  source_grounding      4.36
  domain_placement      4.56
  vsm_relevance         3.31
  explanatory_value     3.94

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

examples/infospace-with-history/output/evaluations/tontines.md

@@ -0,0 +1,64 @@
+---
+entity_slug: tontines
+evaluator: null
+evaluated_at: '2026-02-23T06:32:15.268877'
+overall_score: 4.2
+scores:
+- name: definition_precision
+  value: 5.0
+  max_value: 5.0
+  rationale: The definition is highly precise and non-circular, clearly explaining
+    the specific mechanism of tontines where survivors inherit deceased participants'
+    annuities until one person receives all payments. It captures a distinct financial
+    instrument rather than a vague concept.
+- name: source_grounding
+  value: 5.0
+  max_value: 5.0
+  rationale: This entity is well-grounded in Smith's actual discussion of tontines
+    in Book V, Chapter 3, where he explicitly analyzes them as a government revenue
+    method and explains the psychological factors that make them profitable for governments.
+    The definition accurately reflects Smith's treatment of the topic.
+- name: domain_placement
+  value: 4.0
+  max_value: 5.0
+  rationale: '"Regulation" is appropriate since tontines are a government policy tool
+    for raising revenue, but "Public Finance" or "Government Revenue" might be more
+    precise domains. The placement is reasonable given the regulatory framework required
+    for such schemes.'
+- name: vsm_relevance
+  value: 3.0
+  max_value: 5.0
+  rationale: Tontines map moderately well to S3 (internal regulation) as a government
+    revenue collection mechanism, and potentially to S4 (intelligence) regarding how
+    governments adapt their financing methods. However, the mapping is not as natural
+    as for core operational or regulatory concepts.
+- name: explanatory_value
+  value: 4.0
+  max_value: 5.0
+  rationale: This entity provides good explanatory value by illuminating a specific
+    mechanism of government finance and revealing Smith's insights about how psychological
+    biases (overestimating longevity) can be leveraged for public revenue. It demonstrates
+    concrete policy implementation rather than just naming a surface phenomenon.
+---
+
+# Evaluation: Tontines
+
+## definition_precision — 5.0 / 5.0
+
+The definition is highly precise and non-circular, clearly explaining the specific mechanism of tontines where survivors inherit deceased participants' annuities until one person receives all payments. It captures a distinct financial instrument rather than a vague concept.
+
+## source_grounding — 5.0 / 5.0
+
+This entity is well-grounded in Smith's actual discussion of tontines in Book V, Chapter 3, where he explicitly analyzes them as a government revenue method and explains the psychological factors that make them profitable for governments. The definition accurately reflects Smith's treatment of the topic.
+
+## domain_placement — 4.0 / 5.0
+
+"Regulation" is appropriate since tontines are a government policy tool for raising revenue, but "Public Finance" or "Government Revenue" might be more precise domains. The placement is reasonable given the regulatory framework required for such schemes.
+
+## vsm_relevance — 3.0 / 5.0
+
+Tontines map moderately well to S3 (internal regulation) as a government revenue collection mechanism, and potentially to S4 (intelligence) regarding how governments adapt their financing methods. However, the mapping is not as natural as for core operational or regulatory concepts.
+
+## explanatory_value — 4.0 / 5.0
+
+This entity provides good explanatory value by illuminating a specific mechanism of government finance and revealing Smith's insights about how psychological biases (overestimating longevity) can be leveraged for public revenue. It demonstrates concrete policy implementation rather than just naming a surface phenomenon.