eyeling 1.24.0 → 1.24.1

package/README.md

@@ -14,7 +14,7 @@ echo '@prefix : <http://example.org/> .
 
 ## Read more
 
-- **Handbook:** [eyereasoner.github.io/eyeling/HANDBOOK](https://eyereasoner.github.io/eyeling/HANDBOOK)
-- **Playground:** [eyereasoner.github.io/eyeling/demo](https://eyereasoner.github.io/eyeling/demo)
-- **SEE:** [Specialized Eyeling Executables](see/README.md) — compiled Notation3 examples as standalone JavaScript runners.
-- **Conformance report:** [codeberg.org/phochste/notation3tests/.../report.md](https://codeberg.org/phochste/notation3tests/src/branch/main/reports/report.md)
+- [Handbook](https://eyereasoner.github.io/eyeling/HANDBOOK)
+- [Playground](https://eyereasoner.github.io/eyeling/demo)
+- [SEE: Specialized Eyeling Executables](https://github.com/eyereasoner/eyeling/tree/main/see)
+- [Conformance report](https://codeberg.org/phochste/notation3tests/src/branch/main/reports/report.md)

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "eyeling",
-  "version": "1.24.0",
+  "version": "1.24.1",
   "description": "A minimal Notation3 (N3) reasoner in JavaScript.",
   "main": "./index.js",
   "keywords": [

package/see/README.md

@@ -5,20 +5,19 @@
 SEE (**Specialized Eyeling Executables**) is a small compiler and example suite for turning Eyeling-style Notation3 programs into specialized JavaScript executables with human-readable explanations.
 
 ```text
-input facts -> rules -> trust gate -> insight + explanation
+input facts -> rules -> trust gate -> entailment + explanation
 ```
 
 Each example starts from a Notation3 source and is compiled by `see.js` into a small Node.js program:
 
 - `examples/n3/<name>.n3` contains the N3 facts, rules, gates, and query.
 - `examples/input/<name>.trig` contains the formal RDF 1.2 TriG input evidence read by the runner.
-- `examples/<name>.js` is generated by `see.js` and derives the insight.
+- `examples/<name>.js` is generated by `see.js` and derives the entailment.
 - `examples/output/<name>.md` is the reference Markdown output used by `npm run test:see`.
 - `examples/doc/<name>.md` explains the compiled example.
 
-The trust gate is executable verification. If a required fact is missing, the program fails instead of emitting an unsupported insight.
+The trust gate is executable verification. If a required fact is missing, the program fails instead of emitting an unsupported entailment.
 
-SEE was largely inspired by Prof. Ruben Verborgh's essay [Inside the Insight Economy](https://ruben.verborgh.org/blog/2025/08/12/inside-the-insight-economy/).
 
 ## Run
 
@@ -72,7 +71,7 @@ Validate that the source is inside the supported compiler subset:
 node see/see.js inspect path/to/example.n3
 ```
 
-The compiler currently supports ordinary triples, semicolon/comma property lists, RDF lists, blank nodes, variable predicates, formula-valued conclusions, forward rules, recursive backward rules (`<=`), top-level `log:query`, `=> false` fuses, and a broad native builtin registry for Eyeling-style `math:`, `string:`, `list:`, `crypto:`, `time:`, and `log:` predicates. The compiled JavaScript covers multi-solution predicates such as `list:member`, `list:in`, `list:iterate`, `list:memberAt`, and `list:map`, plus relation-like reverse modes such as `math:exponentiation`, unary trig/degree relations, `list:reverse`, and `list:append` splitting. Offline dereferencing/parsing builtins (`log:content`, `log:semantics`, `log:semanticsOrError`, `log:parsedAsN3`) fail explicitly in compiled SEE examples instead of doing network or parser work at runtime. Unsupported constructs fail rather than jumping directly to a prewritten `log:outputString`. When no authored `log:outputString` report is derived, the compiled runner still emits a SEE-style report with just **Insight** and **Explanation** sections instead of dumping the whole graph; supporting facts are folded into the explanation, including concise support trees that trace selected conclusions back to source facts when the compiler has rule-premise evidence available.
+The compiler currently supports ordinary triples, semicolon/comma property lists, RDF lists, blank nodes, variable predicates, formula-valued conclusions, forward rules, recursive backward rules (`<=`), top-level `log:query`, `=> false` fuses, and a broad native builtin registry for Eyeling-style `math:`, `string:`, `list:`, `crypto:`, `time:`, and `log:` predicates. The compiled JavaScript covers multi-solution predicates such as `list:member`, `list:in`, `list:iterate`, `list:memberAt`, and `list:map`, plus relation-like reverse modes such as `math:exponentiation`, unary trig/degree relations, `list:reverse`, and `list:append` splitting. Offline dereferencing/parsing builtins (`log:content`, `log:semantics`, `log:semanticsOrError`, `log:parsedAsN3`) fail explicitly in compiled SEE examples instead of doing network or parser work at runtime. Unsupported constructs fail rather than jumping directly to a prewritten `log:outputString`. When no authored `log:outputString` report is derived, the compiled runner still emits a SEE-style report with just **Entailment** and **Explanation** sections instead of dumping the whole graph; supporting facts are folded into the explanation, including concise support trees that trace selected entailments back to source facts when the compiler has rule-premise evidence available.
 
 ## Files
 
@@ -81,7 +80,7 @@ examples/
   <name>.js             executable example
   _see.js               shared helper
   input/<name>.trig     formal RDF 1.2 TriG evidence
-  output/<name>.md      reference insight and explanation
+  output/<name>.md      reference entailment and explanation
   doc/<name>.md         guide
   n3/<name>.n3          Notation3 source
 ```
@@ -103,7 +102,7 @@ The repository currently contains **50 N3-compiled SEE examples**. Each example
 | <a id="example-bmi"></a>[BMI](#example-bmi) | compiled from Notation3 by `see.js`; unit normalization, BMI category selection, and healthy-weight range validations are derived from N3 rules. | [doc](examples/doc/bmi.md), [js](examples/bmi.js), [input](examples/input/bmi.trig), [output](examples/output/bmi.md), [n3](examples/n3/bmi.n3) |
 | <a id="example-builtin-coverage-smoke"></a>[Builtin coverage smoke](#example-builtin-coverage-smoke) | compiled from Notation3 by `see.js`; smoke test for compiled math, string, list, crypto, time, and log builtins. | [doc](examples/doc/builtin_coverage.md), [js](examples/builtin_coverage.js), [input](examples/input/builtin_coverage.trig), [output](examples/output/builtin_coverage.md), [n3](examples/n3/builtin_coverage.n3) |
 | <a id="example-collection"></a>[Collection](#example-collection) | compiled from Notation3 by `see.js`; RDF collection/list data with a blank node inside the list. | [doc](examples/doc/collection.md), [js](examples/collection.js), [input](examples/input/collection.trig), [output](examples/output/collection.md), [n3](examples/n3/collection.n3) |
-| <a id="example-complex-numbers"></a>[Complex numbers](#example-complex-numbers) | compiled from Notation3 by `see.js`; complex arithmetic rules and formula-valued conclusion. | [doc](examples/doc/complex.md), [js](examples/complex.js), [input](examples/input/complex.trig), [output](examples/output/complex.md), [n3](examples/n3/complex.n3) |
+| <a id="example-complex-numbers"></a>[Complex numbers](#example-complex-numbers) | compiled from Notation3 by `see.js`; complex arithmetic rules and formula-valued entailment. | [doc](examples/doc/complex.md), [js](examples/complex.js), [input](examples/input/complex.trig), [output](examples/output/complex.md), [n3](examples/n3/complex.n3) |
 | <a id="example-complex-matrix-stability"></a>[Complex matrix stability](#example-complex-matrix-stability) | compiled from Notation3 by `see.js`; diagonal 2x2 complex matrices are classified from squared eigenvalue moduli, with complex product and scaling validations. | [doc](examples/doc/complex_matrix_stability.md), [js](examples/complex_matrix_stability.js), [input](examples/input/complex_matrix_stability.trig), [output](examples/output/complex_matrix_stability.md), [n3](examples/n3/complex_matrix_stability.n3) |
 | <a id="example-composition-of-injective-functions"></a>[Composition of injective functions](#example-composition-of-injective-functions) | compiled from Notation3 by `see.js`; proof-style rules show that the composition of injective functions is injective. | [doc](examples/doc/composition_of_injective_functions_is_injective.md), [js](examples/composition_of_injective_functions_is_injective.js), [input](examples/input/composition_of_injective_functions_is_injective.trig), [output](examples/output/composition_of_injective_functions_is_injective.md), [n3](examples/n3/composition_of_injective_functions_is_injective.n3) |
 | <a id="example-control-system"></a>[Control system](#example-control-system) | compiled from Notation3 by `see.js`; compact EYE-style sensor, disturbance, and target facts derive two actuator commands. | [doc](examples/doc/control_system.md), [js](examples/control_system.js), [input](examples/input/control_system.trig), [output](examples/output/control_system.md), [n3](examples/n3/control_system.n3) |
@@ -111,8 +110,8 @@ The repository currently contains **50 N3-compiled SEE examples**. Each example
 | <a id="example-dijkstra-shortest-path"></a>[Dijkstra shortest path](#example-dijkstra-shortest-path) | compiled from Notation3 by `see.js`; list, collection, sorting, and recursive backward rules derive the shortest path from `:a` to `:f`. | [doc](examples/doc/dijkstra.md), [js](examples/dijkstra.js), [input](examples/input/dijkstra.trig), [output](examples/output/dijkstra.md), [n3](examples/n3/dijkstra.n3) |
 | <a id="example-dijkstra-risk-path"></a>[Dijkstra risk path](#example-dijkstra-risk-path) | compiled from Notation3 by `see.js`; risk-adjusted route selection derives the best path from edge costs, route risk, and threshold rules. | [doc](examples/doc/dijkstra_risk_path.md), [js](examples/dijkstra_risk_path.js), [input](examples/input/dijkstra_risk_path.trig), [output](examples/output/dijkstra_risk_path.md), [n3](examples/n3/dijkstra_risk_path.n3) |
 | <a id="example-dog-license-example"></a>[Dog license example](#example-dog-license-example) | compiled from Notation3 by `see.js`; `log:collectAllIn` and `math:sum` derive that a multi-dog household needs a license. | [doc](examples/doc/dog.md), [js](examples/dog.js), [input](examples/input/dog.trig), [output](examples/output/dog.md), [n3](examples/n3/dog.n3) |
-| <a id="example-eco-route-insight"></a>[Eco route insight](#example-eco-route-insight) | compiled from Notation3 by `see.js`; privacy-preserving route evidence derives an eco-route recommendation with signed-envelope explanation data. | [doc](examples/doc/eco_route_insight.md), [js](examples/eco_route_insight.js), [input](examples/input/eco_route_insight.trig), [output](examples/output/eco_route_insight.md), [n3](examples/n3/eco_route_insight.n3) |
-| <a id="example-delfour"></a>[Delfour](#example-delfour) | compiled from Notation3 by `see.js`; scoped insight-policy facts, cryptographic hash validation, and product rules authorize a lower-sugar scanner recommendation. | [doc](examples/doc/delfour.md), [js](examples/delfour.js), [input](examples/input/delfour.trig), [output](examples/output/delfour.md), [n3](examples/n3/delfour.n3) |
+| <a id="example-eco-route-insight"></a>[Eco route entailment](#example-eco-route-insight) | compiled from Notation3 by `see.js`; privacy-preserving route evidence derives an eco-route recommendation with signed-envelope explanation data. | [doc](examples/doc/eco_route_insight.md), [js](examples/eco_route_insight.js), [input](examples/input/eco_route_insight.trig), [output](examples/output/eco_route_insight.md), [n3](examples/n3/eco_route_insight.n3) |
+| <a id="example-delfour"></a>[Delfour](#example-delfour) | compiled from Notation3 by `see.js`; scoped policy facts, cryptographic hash validation, and product rules authorize a lower-sugar scanner recommendation. | [doc](examples/doc/delfour.md), [js](examples/delfour.js), [input](examples/input/delfour.trig), [output](examples/output/delfour.md), [n3](examples/n3/delfour.n3) |
 | <a id="example-digital-product-passport"></a>[Digital product passport](#example-digital-product-passport) | compiled from Notation3 by `see.js`; component, document, access-policy, and footprint facts fold into a public passport PASS decision. | [doc](examples/doc/digital_product_passport.md), [js](examples/digital_product_passport.js), [input](examples/input/digital_product_passport.trig), [output](examples/output/digital_product_passport.md), [n3](examples/n3/digital_product_passport.n3) |
 | <a id="example-equals-test"></a>[Equals test](#example-equals-test) | compiled from Notation3 by `see.js`; N3 `=` / `owl:sameAs` shorthand compiled into a normal derivation. | [doc](examples/doc/equals.md), [js](examples/equals.js), [input](examples/input/equals.trig), [output](examples/output/equals.md), [n3](examples/n3/equals.n3) |
 | <a id="example-equivalence-class-overlap"></a>[Equivalence class overlap](#example-equivalence-class-overlap) | compiled from Notation3 by `see.js`; equivalence-relation closure proves overlapping equivalence classes are the same class. | [doc](examples/doc/equivalence_classes_overlap_implies_same_class.md), [js](examples/equivalence_classes_overlap_implies_same_class.js), [input](examples/input/equivalence_classes_overlap_implies_same_class.trig), [output](examples/output/equivalence_classes_overlap_implies_same_class.md), [n3](examples/n3/equivalence_classes_overlap_implies_same_class.n3) |
@@ -127,7 +126,7 @@ The repository currently contains **50 N3-compiled SEE examples**. Each example
 | <a id="example-fundamental-theorem-arithmetic"></a>[Fundamental theorem arithmetic](#example-fundamental-theorem-arithmetic) | compiled from Notation3 by `see.js`; primary factorization is validated by product reconstruction, reverse-order sorting, and prime witnesses, with sample summaries retained as data. | [doc](examples/doc/fundamental_theorem_arithmetic.md), [js](examples/fundamental_theorem_arithmetic.js), [input](examples/input/fundamental_theorem_arithmetic.trig), [output](examples/output/fundamental_theorem_arithmetic.md), [n3](examples/n3/fundamental_theorem_arithmetic.n3) |
 | <a id="example-genetic-knapsack-selection"></a>[Genetic knapsack selection](#example-genetic-knapsack-selection) | compiled from Notation3 by `see.js`; deterministic one-bit mutation candidates are scored against knapsack capacity and value constraints. | [doc](examples/doc/genetic_knapsack_selection.md), [js](examples/genetic_knapsack_selection.js), [input](examples/input/genetic_knapsack_selection.trig), [output](examples/output/genetic_knapsack_selection.md), [n3](examples/n3/genetic_knapsack_selection.n3) |
 | <a id="example-goldbach-1000"></a>[Goldbach 1000](#example-goldbach-1000) | compiled from Notation3 by `see.js`; bounded Goldbach run data is validated through representative prime-sum witnesses and a zero-counterexample condition. | [doc](examples/doc/goldbach_1000.md), [js](examples/goldbach_1000.js), [input](examples/input/goldbach_1000.trig), [output](examples/output/goldbach_1000.md), [n3](examples/n3/goldbach_1000.n3) |
-| <a id="example-good-cobbler"></a>[Good cobbler](#example-good-cobbler) | compiled from Notation3 by `see.js`; term-logic-style list pattern and formula-valued conclusion. | [doc](examples/doc/good_cobbler.md), [js](examples/good_cobbler.js), [input](examples/input/good_cobbler.trig), [output](examples/output/good_cobbler.md), [n3](examples/n3/good_cobbler.n3) |
+| <a id="example-good-cobbler"></a>[Good cobbler](#example-good-cobbler) | compiled from Notation3 by `see.js`; term-logic-style list pattern and formula-valued entailment. | [doc](examples/doc/good_cobbler.md), [js](examples/good_cobbler.js), [input](examples/input/good_cobbler.trig), [output](examples/output/good_cobbler.md), [n3](examples/n3/good_cobbler.n3) |
 | <a id="example-gps-route-planning"></a>[GPS route planning](#example-gps-route-planning) | compiled from Notation3 by `see.js`; recursive path rules compare western-Belgium routes by duration, cost, belief, and comfort. | [doc](examples/doc/gps.md), [js](examples/gps.js), [input](examples/input/gps.trig), [output](examples/output/gps.md), [n3](examples/n3/gps.n3) |
 | <a id="example-gray-code-counter"></a>[Gray code counter](#example-gray-code-counter) | compiled from Notation3 by `see.js`; a 4-bit reflected Gray-code cycle is validated as a complete one-bit-transition sequence. | [doc](examples/doc/gray_code_counter.md), [js](examples/gray_code_counter.js), [input](examples/input/gray_code_counter.trig), [output](examples/output/gray_code_counter.md), [n3](examples/n3/gray_code_counter.n3) |
 | <a id="example-greatest-lower-bound-uniqueness"></a>[Greatest lower bound uniqueness](#example-greatest-lower-bound-uniqueness) | compiled from Notation3 by `see.js`; partial-order rules prove two greatest lower bounds of the same pair coincide. | [doc](examples/doc/greatest_lower_bound_uniqueness.md), [js](examples/greatest_lower_bound_uniqueness.js), [input](examples/input/greatest_lower_bound_uniqueness.trig), [output](examples/output/greatest_lower_bound_uniqueness.md), [n3](examples/n3/greatest_lower_bound_uniqueness.n3) |
@@ -143,4 +142,4 @@ The repository currently contains **50 N3-compiled SEE examples**. Each example
 | <a id="example-smoke-arithmetic"></a>[Smoke Arithmetic](#example-smoke-arithmetic) | compiled from Notation3 by `see.js`; arithmetic compiler smoke test deriving 6 × 7 through rules and builtins. | [doc](examples/doc/smoke_arithmetic.md), [js](examples/smoke_arithmetic.js), [input](examples/input/smoke_arithmetic.trig), [output](examples/output/smoke_arithmetic.md), [n3](examples/n3/smoke_arithmetic.n3) |
 | <a id="example-socrates-inference"></a>[Socrates inference](#example-socrates-inference) | compiled from Notation3 by `see.js`; subclass closure with top-level `log:query` projection. | [doc](examples/doc/socrates.md), [js](examples/socrates.js), [input](examples/input/socrates.trig), [output](examples/output/socrates.md), [n3](examples/n3/socrates.n3) |
 | <a id="example-wind-turbine-envelope"></a>[Wind turbine envelope](#example-wind-turbine-envelope) | compiled from Notation3 by `see.js`; wind-speed samples are classified against cut-in, rated, and cut-out thresholds and accumulated into energy. | [doc](examples/doc/wind_turbine.md), [js](examples/wind_turbine.js), [input](examples/input/wind_turbine.trig), [output](examples/output/wind_turbine.md), [n3](examples/n3/wind_turbine.n3) |
-| <a id="example-burn-the-witch"></a>[Burn the witch](#example-burn-the-witch) | compiled from Notation3 by `see.js`; rule-chain explanation for the witch conclusion. | [doc](examples/doc/witch.md), [js](examples/witch.js), [input](examples/input/witch.trig), [output](examples/output/witch.md), [n3](examples/n3/witch.n3) |
+| <a id="example-burn-the-witch"></a>[Burn the witch](#example-burn-the-witch) | compiled from Notation3 by `see.js`; rule-chain explanation for the witch entailment. | [doc](examples/doc/witch.md), [js](examples/witch.js), [input](examples/input/witch.trig), [output](examples/output/witch.md), [n3](examples/n3/witch.n3) |

package/see/examples/age.js

@@ -1040,19 +1040,19 @@ function renderStructuredOutput({ title, graph, queries = [], rules = [], initia
   const lines = [];
   lines.push('# ' + title);
   lines.push('');
-  lines.push('## Insight');
+  lines.push('## Entailment');
   if (mode === 'query') {
     lines.push('The compiled query selected ' + selected.length + ' fact(s) after the rule closure was computed.');
   } else if (mode === 'formula') {
-    lines.push('The derivation produced ' + selected.length + ' formula-valued conclusion(s).');
+    lines.push('The derivation produced ' + selected.length + ' formula-valued entailment(s).');
   } else {
     lines.push('The derivation produced ' + derived.length + ' new fact(s) from ' + initialFacts.length + ' stated fact(s).');
   }
-  if (keyFact) lines.push('Main conclusion: **' + factSentence(keyFact) + '**');
+  if (keyFact) lines.push('Main entailment: **' + factSentence(keyFact) + '**');
   const bullets = selected.slice(-6).reverse();
   if (bullets.length) {
     lines.push('');
-    lines.push('Selected conclusions:');
+    lines.push('Selected entailments:');
     for (const fact of bullets) lines.push('- ' + codeFact(fact));
   }
   lines.push('');
@@ -1114,15 +1114,15 @@ function dedupeExplanationHeadings(text) {
 function normalizePublicReport(markdown, title) {
   let text = String(markdown || '').trimEnd();
   if (!/^\s*#\s+/m.test(text)) text = '# ' + title + '\n\n' + text;
-  if (!/^##\s+Insight\s*$/mi.test(text)) {
-    text = text.replace(/^(#\s+[^\n]+\n*)/, '$1\n## Insight\n');
+  if (!/^##\s+Entailment\s*$/mi.test(text)) {
+    text = text.replace(/^(#\s+[^\n]+\n*)/, '$1\n## Entailment\n');
   }
   if (!/^##\s+Explanation\s*$/mi.test(text)) {
     text += '\n\n## Explanation\nNo additional explanation was provided by the generated output.';
   }
   text = text.replace(/^##\s+([^\n]+?)\s*$/gm, (line, heading) => {
     const normalized = heading.trim().toLowerCase();
-    if (normalized === 'insight' || normalized === 'explanation') return '## ' + (normalized === 'insight' ? 'Insight' : 'Explanation');
+    if (normalized === 'insight' || normalized === 'conclusion' || normalized === 'entailment' || normalized === 'explanation') return '## ' + (normalized === 'explanation' ? 'Explanation' : 'Entailment');
     return '**' + heading.trim() + '**';
   });
   text = dedupeExplanationHeadings(text);
@@ -1131,13 +1131,13 @@ function normalizePublicReport(markdown, title) {
 function markdownize(raw, title) {
   let text = String(raw || '');
   text = text
-    .replace(/===\s*Answer\s*===/g, '## Insight')
+    .replace(/===\s*Answer\s*===/g, '## Entailment')
     .replace(/===\s*Reason\s+Why\s*===/gi, '## Explanation')
     .replace(/===\s*Explanation\s*===/gi, '## Explanation')
     .replace(/===\s*([^=]+?)\s*===/g, (_, h) => '**' + h.trim() + '**');
   text = text.replace(/^C(\d+)\s+OK\s*-\s*/gm, 'C$1: ');
   text = dedupeExplanationHeadings(text);
-  if (!text.trim()) text = '## Insight\nNo log:outputString facts were derived.\n\n## Explanation\nThe compiled derivation did not produce authored report text.';
+  if (!text.trim()) text = '## Entailment\nNo log:outputString facts were derived.\n\n## Explanation\nThe compiled derivation did not produce authored report text.';
   return normalizePublicReport(text, title);
 }
 function authoredSupportAppendix(graph, queries, rules, initialFacts, trace) {
@@ -1334,7 +1334,7 @@ const QUERIES = [
     ]
   }
 ];
-const DOC_MARKDOWN = "# Age checker\n\nGenerated by `see.js` from a Notation3 source file.\n\nIs the age of a person above some duration?\n\n## Compilation summary\n\n- Example name: `age`\n- Input facts emitted: 1\n- Forward rules compiled: 0\n- Backward predicate rules compiled: 1\n- Fuses compiled: 0\n- Predicate count: 6\n\n## Built-ins used\n\n- `math:difference`\n- `math:greaterThan`\n\n## Runtime model\n\nThe generated `examples/age.js` is a specialized JavaScript derivation program. For ordinary sources, `see.js` emits the source facts as `examples/input/age.trig`. For rules-only sources, generation can reuse an existing external evidence file such as `examples/input/age.trig` or `examples/input/age.trig`. The runner reads that TriG evidence directly and performs a local fixpoint derivation; it does not parse the program source or call an external reasoner.\n\n## Output model\n\nRunning `node examples/age.js` produces a SEE-style Markdown report with an **Insight** section, an **Explanation** section, and a **Formal TriG Output** section containing the selected derived/query facts.\n";
+const DOC_MARKDOWN = "# Age checker\n\nGenerated by `see.js` from a Notation3 source file.\n\nIs the age of a person above some duration?\n\n## Compilation summary\n\n- Example name: `age`\n- Input facts emitted: 1\n- Forward rules compiled: 0\n- Backward predicate rules compiled: 1\n- Fuses compiled: 0\n- Predicate count: 6\n\n## Built-ins used\n\n- `math:difference`\n- `math:greaterThan`\n\n## Runtime model\n\nThe generated `examples/age.js` is a specialized JavaScript derivation program. For ordinary sources, `see.js` emits the source facts as `examples/input/age.trig`. For rules-only sources, generation can reuse an existing external evidence file such as `examples/input/age.trig` or `examples/input/age.trig`. The runner reads that TriG evidence directly and performs a local fixpoint derivation; it does not parse the program source or call an external reasoner.\n\n## Output model\n\nRunning `node examples/age.js` produces a SEE-style Markdown report with an **Entailment** section, an **Explanation** section, and a **Formal TriG Output** section containing the selected derived/query facts.\n";
 function seeMetadata(data) { return (data && data.__see) || {}; }
 function trustedDerivation(data) { const meta = seeMetadata(data); const facts = data && Array.isArray(data.facts) ? data.facts : []; const expectedFacts = EXPECTED_INPUT_FACTS || Number(meta.InputFacts || 0); if (meta.SourceSHA256 && meta.SourceSHA256 !== "bf52bb918b595bb68a0a6304729e8129f3a851dc16cd3980ec5932115f2e7aea") throw new Error('input evidence does not match the N3 source compiled into this example'); const result = saturate(facts, RULES); const rawOutput = renderRawOutput(result.graph, QUERIES, RULES, facts); fail('Compiled N3 derivation failed', { 'input evidence metadata is present and matches compiled source': meta.SourceSHA256 === "bf52bb918b595bb68a0a6304729e8129f3a851dc16cd3980ec5932115f2e7aea", 'input evidence facts were loaded': expectedFacts > 0 ? facts.length === expectedFacts : facts.length >= 0, 'compiled rules were loaded': RULES.length === 1, 'compiled query directives were loaded': QUERIES.length === 1, 'a derivation fixpoint was reached': result.graph.facts.length >= facts.length, 'query or output facts were produced': rawOutput.length > 0 }); return { ...result, rawOutput, inputFacts: facts }; }
 function snapshotMarkdown(markdown) { return markdown.split(/\n/).map((line) => line ? line + '  \n' : '\n').join(''); }

package/see/examples/annotation.js

@@ -1040,19 +1040,19 @@ function renderStructuredOutput({ title, graph, queries = [], rules = [], initia
   const lines = [];
   lines.push('# ' + title);
   lines.push('');
-  lines.push('## Insight');
+  lines.push('## Entailment');
   if (mode === 'query') {
     lines.push('The compiled query selected ' + selected.length + ' fact(s) after the rule closure was computed.');
   } else if (mode === 'formula') {
-    lines.push('The derivation produced ' + selected.length + ' formula-valued conclusion(s).');
+    lines.push('The derivation produced ' + selected.length + ' formula-valued entailment(s).');
   } else {
     lines.push('The derivation produced ' + derived.length + ' new fact(s) from ' + initialFacts.length + ' stated fact(s).');
   }
-  if (keyFact) lines.push('Main conclusion: **' + factSentence(keyFact) + '**');
+  if (keyFact) lines.push('Main entailment: **' + factSentence(keyFact) + '**');
   const bullets = selected.slice(-6).reverse();
   if (bullets.length) {
     lines.push('');
-    lines.push('Selected conclusions:');
+    lines.push('Selected entailments:');
     for (const fact of bullets) lines.push('- ' + codeFact(fact));
   }
   lines.push('');
@@ -1114,15 +1114,15 @@ function dedupeExplanationHeadings(text) {
 function normalizePublicReport(markdown, title) {
   let text = String(markdown || '').trimEnd();
   if (!/^\s*#\s+/m.test(text)) text = '# ' + title + '\n\n' + text;
-  if (!/^##\s+Insight\s*$/mi.test(text)) {
-    text = text.replace(/^(#\s+[^\n]+\n*)/, '$1\n## Insight\n');
+  if (!/^##\s+Entailment\s*$/mi.test(text)) {
+    text = text.replace(/^(#\s+[^\n]+\n*)/, '$1\n## Entailment\n');
   }
   if (!/^##\s+Explanation\s*$/mi.test(text)) {
     text += '\n\n## Explanation\nNo additional explanation was provided by the generated output.';
   }
   text = text.replace(/^##\s+([^\n]+?)\s*$/gm, (line, heading) => {
     const normalized = heading.trim().toLowerCase();
-    if (normalized === 'insight' || normalized === 'explanation') return '## ' + (normalized === 'insight' ? 'Insight' : 'Explanation');
+    if (normalized === 'insight' || normalized === 'conclusion' || normalized === 'entailment' || normalized === 'explanation') return '## ' + (normalized === 'explanation' ? 'Explanation' : 'Entailment');
     return '**' + heading.trim() + '**';
   });
   text = dedupeExplanationHeadings(text);
@@ -1131,13 +1131,13 @@ function normalizePublicReport(markdown, title) {
 function markdownize(raw, title) {
   let text = String(raw || '');
   text = text
-    .replace(/===\s*Answer\s*===/g, '## Insight')
+    .replace(/===\s*Answer\s*===/g, '## Entailment')
     .replace(/===\s*Reason\s+Why\s*===/gi, '## Explanation')
     .replace(/===\s*Explanation\s*===/gi, '## Explanation')
     .replace(/===\s*([^=]+?)\s*===/g, (_, h) => '**' + h.trim() + '**');
   text = text.replace(/^C(\d+)\s+OK\s*-\s*/gm, 'C$1: ');
   text = dedupeExplanationHeadings(text);
-  if (!text.trim()) text = '## Insight\nNo log:outputString facts were derived.\n\n## Explanation\nThe compiled derivation did not produce authored report text.';
+  if (!text.trim()) text = '## Entailment\nNo log:outputString facts were derived.\n\n## Explanation\nThe compiled derivation did not produce authored report text.';
   return normalizePublicReport(text, title);
 }
 function authoredSupportAppendix(graph, queries, rules, initialFacts, trace) {
@@ -1195,7 +1195,7 @@ const TITLE = "Annotation";
 const EXPECTED_INPUT_FACTS = 4;
 const RULES = [];
 const QUERIES = [];
-const DOC_MARKDOWN = "# Annotation\n\nGenerated by `see.js` from a Notation3 source file.\n\n## Compilation summary\n\n- Example name: `annotation`\n- Input facts emitted: 4\n- Forward rules compiled: 0\n- Backward predicate rules compiled: 0\n- Fuses compiled: 0\n- Predicate count: 4\n\n## Built-ins used\n\n- `log:nameOf`\n\n## Runtime model\n\nThe generated `examples/annotation.js` is a specialized JavaScript derivation program. For ordinary sources, `see.js` emits the source facts as `examples/input/annotation.trig`. For rules-only sources, generation can reuse an existing external evidence file such as `examples/input/annotation.trig` or `examples/input/annotation.trig`. The runner reads that TriG evidence directly and performs a local fixpoint derivation; it does not parse the program source or call an external reasoner.\n\n## Output model\n\nRunning `node examples/annotation.js` produces a SEE-style Markdown report with an **Insight** section, an **Explanation** section, and a **Formal TriG Output** section containing the selected derived/query facts.\n";
+const DOC_MARKDOWN = "# Annotation\n\nGenerated by `see.js` from a Notation3 source file.\n\n## Compilation summary\n\n- Example name: `annotation`\n- Input facts emitted: 4\n- Forward rules compiled: 0\n- Backward predicate rules compiled: 0\n- Fuses compiled: 0\n- Predicate count: 4\n\n## Built-ins used\n\n- `log:nameOf`\n\n## Runtime model\n\nThe generated `examples/annotation.js` is a specialized JavaScript derivation program. For ordinary sources, `see.js` emits the source facts as `examples/input/annotation.trig`. For rules-only sources, generation can reuse an existing external evidence file such as `examples/input/annotation.trig` or `examples/input/annotation.trig`. The runner reads that TriG evidence directly and performs a local fixpoint derivation; it does not parse the program source or call an external reasoner.\n\n## Output model\n\nRunning `node examples/annotation.js` produces a SEE-style Markdown report with an **Entailment** section, an **Explanation** section, and a **Formal TriG Output** section containing the selected derived/query facts.\n";
 function seeMetadata(data) { return (data && data.__see) || {}; }
 function trustedDerivation(data) { const meta = seeMetadata(data); const facts = data && Array.isArray(data.facts) ? data.facts : []; const expectedFacts = EXPECTED_INPUT_FACTS || Number(meta.InputFacts || 0); if (meta.SourceSHA256 && meta.SourceSHA256 !== "fa6faef8fac2e328ccfb894c21eef260b0935a936b5e9806f8889719e7f39e4f") throw new Error('input evidence does not match the N3 source compiled into this example'); const result = saturate(facts, RULES); const rawOutput = renderRawOutput(result.graph, QUERIES, RULES, facts); fail('Compiled N3 derivation failed', { 'input evidence metadata is present and matches compiled source': meta.SourceSHA256 === "fa6faef8fac2e328ccfb894c21eef260b0935a936b5e9806f8889719e7f39e4f", 'input evidence facts were loaded': expectedFacts > 0 ? facts.length === expectedFacts : facts.length >= 0, 'compiled rules were loaded': RULES.length === 0, 'compiled query directives were loaded': QUERIES.length === 0, 'a derivation fixpoint was reached': result.graph.facts.length >= facts.length, 'query or output facts were produced': rawOutput.length > 0 }); return { ...result, rawOutput, inputFacts: facts }; }
 function snapshotMarkdown(markdown) { return markdown.split(/\n/).map((line) => line ? line + '  \n' : '\n').join(''); }

package/see/examples/backward.js

@@ -1040,19 +1040,19 @@ function renderStructuredOutput({ title, graph, queries = [], rules = [], initia
   const lines = [];
   lines.push('# ' + title);
   lines.push('');
-  lines.push('## Insight');
+  lines.push('## Entailment');
   if (mode === 'query') {
     lines.push('The compiled query selected ' + selected.length + ' fact(s) after the rule closure was computed.');
   } else if (mode === 'formula') {
-    lines.push('The derivation produced ' + selected.length + ' formula-valued conclusion(s).');
+    lines.push('The derivation produced ' + selected.length + ' formula-valued entailment(s).');
   } else {
     lines.push('The derivation produced ' + derived.length + ' new fact(s) from ' + initialFacts.length + ' stated fact(s).');
   }
-  if (keyFact) lines.push('Main conclusion: **' + factSentence(keyFact) + '**');
+  if (keyFact) lines.push('Main entailment: **' + factSentence(keyFact) + '**');
   const bullets = selected.slice(-6).reverse();
   if (bullets.length) {
     lines.push('');
-    lines.push('Selected conclusions:');
+    lines.push('Selected entailments:');
     for (const fact of bullets) lines.push('- ' + codeFact(fact));
   }
   lines.push('');
@@ -1114,15 +1114,15 @@ function dedupeExplanationHeadings(text) {
 function normalizePublicReport(markdown, title) {
   let text = String(markdown || '').trimEnd();
   if (!/^\s*#\s+/m.test(text)) text = '# ' + title + '\n\n' + text;
-  if (!/^##\s+Insight\s*$/mi.test(text)) {
-    text = text.replace(/^(#\s+[^\n]+\n*)/, '$1\n## Insight\n');
+  if (!/^##\s+Entailment\s*$/mi.test(text)) {
+    text = text.replace(/^(#\s+[^\n]+\n*)/, '$1\n## Entailment\n');
   }
   if (!/^##\s+Explanation\s*$/mi.test(text)) {
     text += '\n\n## Explanation\nNo additional explanation was provided by the generated output.';
   }
   text = text.replace(/^##\s+([^\n]+?)\s*$/gm, (line, heading) => {
     const normalized = heading.trim().toLowerCase();
-    if (normalized === 'insight' || normalized === 'explanation') return '## ' + (normalized === 'insight' ? 'Insight' : 'Explanation');
+    if (normalized === 'insight' || normalized === 'conclusion' || normalized === 'entailment' || normalized === 'explanation') return '## ' + (normalized === 'explanation' ? 'Explanation' : 'Entailment');
     return '**' + heading.trim() + '**';
   });
   text = dedupeExplanationHeadings(text);
@@ -1131,13 +1131,13 @@ function normalizePublicReport(markdown, title) {
 function markdownize(raw, title) {
   let text = String(raw || '');
   text = text
-    .replace(/===\s*Answer\s*===/g, '## Insight')
+    .replace(/===\s*Answer\s*===/g, '## Entailment')
     .replace(/===\s*Reason\s+Why\s*===/gi, '## Explanation')
     .replace(/===\s*Explanation\s*===/gi, '## Explanation')
     .replace(/===\s*([^=]+?)\s*===/g, (_, h) => '**' + h.trim() + '**');
   text = text.replace(/^C(\d+)\s+OK\s*-\s*/gm, 'C$1: ');
   text = dedupeExplanationHeadings(text);
-  if (!text.trim()) text = '## Insight\nNo log:outputString facts were derived.\n\n## Explanation\nThe compiled derivation did not produce authored report text.';
+  if (!text.trim()) text = '## Entailment\nNo log:outputString facts were derived.\n\n## Explanation\nThe compiled derivation did not produce authored report text.';
   return normalizePublicReport(text, title);
 }
 function authoredSupportAppendix(graph, queries, rules, initialFacts, trace) {
@@ -1280,7 +1280,7 @@ const RULES = [
   }
 ];
 const QUERIES = [];
-const DOC_MARKDOWN = "# Backward rule example\n\nGenerated by `see.js` from a Notation3 source file.\n\nSee https://www.w3.org/2000/10/swap/doc/tutorial-1.pdf page 17\n\n## Compilation summary\n\n- Example name: `backward`\n- Input facts emitted: 0\n- Forward rules compiled: 1\n- Backward predicate rules compiled: 1\n- Fuses compiled: 0\n- Predicate count: 3\n\n## Built-ins used\n\n- `math:greaterThan`\n\n## Runtime model\n\nThe generated `examples/backward.js` is a specialized JavaScript derivation program. For ordinary sources, `see.js` emits the source facts as `examples/input/backward.trig`. For rules-only sources, generation can reuse an existing external evidence file such as `examples/input/backward.trig` or `examples/input/backward.trig`. The runner reads that TriG evidence directly and performs a local fixpoint derivation; it does not parse the program source or call an external reasoner.\n\n## Output model\n\nRunning `node examples/backward.js` produces a SEE-style Markdown report with an **Insight** section, an **Explanation** section, and a **Formal TriG Output** section containing the selected derived/query facts.\n";
+const DOC_MARKDOWN = "# Backward rule example\n\nGenerated by `see.js` from a Notation3 source file.\n\nSee https://www.w3.org/2000/10/swap/doc/tutorial-1.pdf page 17\n\n## Compilation summary\n\n- Example name: `backward`\n- Input facts emitted: 0\n- Forward rules compiled: 1\n- Backward predicate rules compiled: 1\n- Fuses compiled: 0\n- Predicate count: 3\n\n## Built-ins used\n\n- `math:greaterThan`\n\n## Runtime model\n\nThe generated `examples/backward.js` is a specialized JavaScript derivation program. For ordinary sources, `see.js` emits the source facts as `examples/input/backward.trig`. For rules-only sources, generation can reuse an existing external evidence file such as `examples/input/backward.trig` or `examples/input/backward.trig`. The runner reads that TriG evidence directly and performs a local fixpoint derivation; it does not parse the program source or call an external reasoner.\n\n## Output model\n\nRunning `node examples/backward.js` produces a SEE-style Markdown report with an **Entailment** section, an **Explanation** section, and a **Formal TriG Output** section containing the selected derived/query facts.\n";
 function seeMetadata(data) { return (data && data.__see) || {}; }
 function trustedDerivation(data) { const meta = seeMetadata(data); const facts = data && Array.isArray(data.facts) ? data.facts : []; const expectedFacts = EXPECTED_INPUT_FACTS || Number(meta.InputFacts || 0); if (meta.SourceSHA256 && meta.SourceSHA256 !== "328d66402900d98dda0c02e3c64a7f3633f01a0ba148031c17e8fbf353057d33") throw new Error('input evidence does not match the N3 source compiled into this example'); const result = saturate(facts, RULES); const rawOutput = renderRawOutput(result.graph, QUERIES, RULES, facts); fail('Compiled N3 derivation failed', { 'input evidence metadata is present and matches compiled source': meta.SourceSHA256 === "328d66402900d98dda0c02e3c64a7f3633f01a0ba148031c17e8fbf353057d33", 'input evidence facts were loaded': expectedFacts > 0 ? facts.length === expectedFacts : facts.length >= 0, 'compiled rules were loaded': RULES.length === 2, 'compiled query directives were loaded': QUERIES.length === 0, 'a derivation fixpoint was reached': result.graph.facts.length >= facts.length, 'query or output facts were produced': rawOutput.length > 0 }); return { ...result, rawOutput, inputFacts: facts }; }
 function snapshotMarkdown(markdown) { return markdown.split(/\n/).map((line) => line ? line + '  \n' : '\n').join(''); }

package/see/examples/backward_recursion.js

@@ -1040,19 +1040,19 @@ function renderStructuredOutput({ title, graph, queries = [], rules = [], initia
   const lines = [];
   lines.push('# ' + title);
   lines.push('');
-  lines.push('## Insight');
+  lines.push('## Entailment');
   if (mode === 'query') {
     lines.push('The compiled query selected ' + selected.length + ' fact(s) after the rule closure was computed.');
   } else if (mode === 'formula') {
-    lines.push('The derivation produced ' + selected.length + ' formula-valued conclusion(s).');
+    lines.push('The derivation produced ' + selected.length + ' formula-valued entailment(s).');
   } else {
     lines.push('The derivation produced ' + derived.length + ' new fact(s) from ' + initialFacts.length + ' stated fact(s).');
   }
-  if (keyFact) lines.push('Main conclusion: **' + factSentence(keyFact) + '**');
+  if (keyFact) lines.push('Main entailment: **' + factSentence(keyFact) + '**');
   const bullets = selected.slice(-6).reverse();
   if (bullets.length) {
     lines.push('');
-    lines.push('Selected conclusions:');
+    lines.push('Selected entailments:');
     for (const fact of bullets) lines.push('- ' + codeFact(fact));
   }
   lines.push('');
@@ -1114,15 +1114,15 @@ function dedupeExplanationHeadings(text) {
 function normalizePublicReport(markdown, title) {
   let text = String(markdown || '').trimEnd();
   if (!/^\s*#\s+/m.test(text)) text = '# ' + title + '\n\n' + text;
-  if (!/^##\s+Insight\s*$/mi.test(text)) {
-    text = text.replace(/^(#\s+[^\n]+\n*)/, '$1\n## Insight\n');
+  if (!/^##\s+Entailment\s*$/mi.test(text)) {
+    text = text.replace(/^(#\s+[^\n]+\n*)/, '$1\n## Entailment\n');
   }
   if (!/^##\s+Explanation\s*$/mi.test(text)) {
     text += '\n\n## Explanation\nNo additional explanation was provided by the generated output.';
   }
   text = text.replace(/^##\s+([^\n]+?)\s*$/gm, (line, heading) => {
     const normalized = heading.trim().toLowerCase();
-    if (normalized === 'insight' || normalized === 'explanation') return '## ' + (normalized === 'insight' ? 'Insight' : 'Explanation');
+    if (normalized === 'insight' || normalized === 'conclusion' || normalized === 'entailment' || normalized === 'explanation') return '## ' + (normalized === 'explanation' ? 'Explanation' : 'Entailment');
     return '**' + heading.trim() + '**';
   });
   text = dedupeExplanationHeadings(text);
@@ -1131,13 +1131,13 @@ function normalizePublicReport(markdown, title) {
 function markdownize(raw, title) {
   let text = String(raw || '');
   text = text
-    .replace(/===\s*Answer\s*===/g, '## Insight')
+    .replace(/===\s*Answer\s*===/g, '## Entailment')
     .replace(/===\s*Reason\s+Why\s*===/gi, '## Explanation')
     .replace(/===\s*Explanation\s*===/gi, '## Explanation')
     .replace(/===\s*([^=]+?)\s*===/g, (_, h) => '**' + h.trim() + '**');
   text = text.replace(/^C(\d+)\s+OK\s*-\s*/gm, 'C$1: ');
   text = dedupeExplanationHeadings(text);
-  if (!text.trim()) text = '## Insight\nNo log:outputString facts were derived.\n\n## Explanation\nThe compiled derivation did not produce authored report text.';
+  if (!text.trim()) text = '## Entailment\nNo log:outputString facts were derived.\n\n## Explanation\nThe compiled derivation did not produce authored report text.';
   return normalizePublicReport(text, title);
 }
 function authoredSupportAppendix(graph, queries, rules, initialFacts, trace) {
@@ -1379,7 +1379,7 @@ const QUERIES = [
     ]
   }
 ];
-const DOC_MARKDOWN = "# Backward recursion coverage\n\nGenerated by `see.js` from a Notation3 source file.\n\nDemonstrates recursive <= rules compiled into specialized JavaScript.\n\n## Compilation summary\n\n- Example name: `backward_recursion`\n- Input facts emitted: 2\n- Forward rules compiled: 1\n- Backward predicate rules compiled: 2\n- Fuses compiled: 0\n- Predicate count: 3\n\n## Built-ins used\n\n- none\n\n## Runtime model\n\nThe generated `examples/backward_recursion.js` is a specialized JavaScript derivation program. For ordinary sources, `see.js` emits the source facts as `examples/input/backward_recursion.trig`. For rules-only sources, generation can reuse an existing external evidence file such as `examples/input/backward-recursion.trig` or `examples/input/backward_recursion.trig`. The runner reads that TriG evidence directly and performs a local fixpoint derivation; it does not parse the program source or call an external reasoner.\n\n## Output model\n\nRunning `node examples/backward_recursion.js` produces a SEE-style Markdown report with an **Insight** section, an **Explanation** section, and a **Formal TriG Output** section containing the selected derived/query facts.\n";
+const DOC_MARKDOWN = "# Backward recursion coverage\n\nGenerated by `see.js` from a Notation3 source file.\n\nDemonstrates recursive <= rules compiled into specialized JavaScript.\n\n## Compilation summary\n\n- Example name: `backward_recursion`\n- Input facts emitted: 2\n- Forward rules compiled: 1\n- Backward predicate rules compiled: 2\n- Fuses compiled: 0\n- Predicate count: 3\n\n## Built-ins used\n\n- none\n\n## Runtime model\n\nThe generated `examples/backward_recursion.js` is a specialized JavaScript derivation program. For ordinary sources, `see.js` emits the source facts as `examples/input/backward_recursion.trig`. For rules-only sources, generation can reuse an existing external evidence file such as `examples/input/backward-recursion.trig` or `examples/input/backward_recursion.trig`. The runner reads that TriG evidence directly and performs a local fixpoint derivation; it does not parse the program source or call an external reasoner.\n\n## Output model\n\nRunning `node examples/backward_recursion.js` produces a SEE-style Markdown report with an **Entailment** section, an **Explanation** section, and a **Formal TriG Output** section containing the selected derived/query facts.\n";
 function seeMetadata(data) { return (data && data.__see) || {}; }
 function trustedDerivation(data) { const meta = seeMetadata(data); const facts = data && Array.isArray(data.facts) ? data.facts : []; const expectedFacts = EXPECTED_INPUT_FACTS || Number(meta.InputFacts || 0); if (meta.SourceSHA256 && meta.SourceSHA256 !== "1249b9825793b7f0174c87a4d0e44752cb63f3940cc332c0d00fc6bb15a081d1") throw new Error('input evidence does not match the N3 source compiled into this example'); const result = saturate(facts, RULES); const rawOutput = renderRawOutput(result.graph, QUERIES, RULES, facts); fail('Compiled N3 derivation failed', { 'input evidence metadata is present and matches compiled source': meta.SourceSHA256 === "1249b9825793b7f0174c87a4d0e44752cb63f3940cc332c0d00fc6bb15a081d1", 'input evidence facts were loaded': expectedFacts > 0 ? facts.length === expectedFacts : facts.length >= 0, 'compiled rules were loaded': RULES.length === 3, 'compiled query directives were loaded': QUERIES.length === 1, 'a derivation fixpoint was reached': result.graph.facts.length >= facts.length, 'query or output facts were produced': rawOutput.length > 0 }); return { ...result, rawOutput, inputFacts: facts }; }
 function snapshotMarkdown(markdown) { return markdown.split(/\n/).map((line) => line ? line + '  \n' : '\n').join(''); }

package/see/examples/bayes_diagnosis.js

@@ -1040,19 +1040,19 @@ function renderStructuredOutput({ title, graph, queries = [], rules = [], initia
   const lines = [];
   lines.push('# ' + title);
   lines.push('');
-  lines.push('## Insight');
+  lines.push('## Entailment');
   if (mode === 'query') {
     lines.push('The compiled query selected ' + selected.length + ' fact(s) after the rule closure was computed.');
   } else if (mode === 'formula') {
-    lines.push('The derivation produced ' + selected.length + ' formula-valued conclusion(s).');
+    lines.push('The derivation produced ' + selected.length + ' formula-valued entailment(s).');
   } else {
     lines.push('The derivation produced ' + derived.length + ' new fact(s) from ' + initialFacts.length + ' stated fact(s).');
   }
-  if (keyFact) lines.push('Main conclusion: **' + factSentence(keyFact) + '**');
+  if (keyFact) lines.push('Main entailment: **' + factSentence(keyFact) + '**');
   const bullets = selected.slice(-6).reverse();
   if (bullets.length) {
     lines.push('');
-    lines.push('Selected conclusions:');
+    lines.push('Selected entailments:');
     for (const fact of bullets) lines.push('- ' + codeFact(fact));
   }
   lines.push('');
@@ -1114,15 +1114,15 @@ function dedupeExplanationHeadings(text) {
 function normalizePublicReport(markdown, title) {
   let text = String(markdown || '').trimEnd();
   if (!/^\s*#\s+/m.test(text)) text = '# ' + title + '\n\n' + text;
-  if (!/^##\s+Insight\s*$/mi.test(text)) {
-    text = text.replace(/^(#\s+[^\n]+\n*)/, '$1\n## Insight\n');
+  if (!/^##\s+Entailment\s*$/mi.test(text)) {
+    text = text.replace(/^(#\s+[^\n]+\n*)/, '$1\n## Entailment\n');
   }
   if (!/^##\s+Explanation\s*$/mi.test(text)) {
     text += '\n\n## Explanation\nNo additional explanation was provided by the generated output.';
   }
   text = text.replace(/^##\s+([^\n]+?)\s*$/gm, (line, heading) => {
     const normalized = heading.trim().toLowerCase();
-    if (normalized === 'insight' || normalized === 'explanation') return '## ' + (normalized === 'insight' ? 'Insight' : 'Explanation');
+    if (normalized === 'insight' || normalized === 'conclusion' || normalized === 'entailment' || normalized === 'explanation') return '## ' + (normalized === 'explanation' ? 'Explanation' : 'Entailment');
     return '**' + heading.trim() + '**';
   });
   text = dedupeExplanationHeadings(text);
@@ -1131,13 +1131,13 @@ function normalizePublicReport(markdown, title) {
 function markdownize(raw, title) {
   let text = String(raw || '');
   text = text
-    .replace(/===\s*Answer\s*===/g, '## Insight')
+    .replace(/===\s*Answer\s*===/g, '## Entailment')
     .replace(/===\s*Reason\s+Why\s*===/gi, '## Explanation')
     .replace(/===\s*Explanation\s*===/gi, '## Explanation')
     .replace(/===\s*([^=]+?)\s*===/g, (_, h) => '**' + h.trim() + '**');
   text = text.replace(/^C(\d+)\s+OK\s*-\s*/gm, 'C$1: ');
   text = dedupeExplanationHeadings(text);
-  if (!text.trim()) text = '## Insight\nNo log:outputString facts were derived.\n\n## Explanation\nThe compiled derivation did not produce authored report text.';
+  if (!text.trim()) text = '## Entailment\nNo log:outputString facts were derived.\n\n## Explanation\nThe compiled derivation did not produce authored report text.';
   return normalizePublicReport(text, title);
 }
 function authoredSupportAppendix(graph, queries, rules, initialFacts, trace) {
@@ -2758,7 +2758,7 @@ const QUERIES = [
     ]
   }
 ];
-const DOC_MARKDOWN = "# Bayes Diagnosis\n\nGenerated by `see.js` from a Notation3 source file.\n\nNaive Bayes diagnosis example, adapted from Eyeling's bayes-diagnosis.n3.\nThe facts and rules compute unnormalized likelihoods, normalize them into\nposterior probabilities, and emit a SEE report for the highest posterior.\nValues are illustrative only and are not medical advice.\n\n## Compilation summary\n\n- Example name: `bayes_diagnosis`\n- Input facts emitted: 85\n- Forward rules compiled: 4\n- Backward predicate rules compiled: 5\n- Fuses compiled: 4\n- Predicate count: 34\n\n## Built-ins used\n\n- `list:firstRest`\n- `list:iterate`\n- `list:map`\n- `list:memberAt`\n- `log:outputString`\n- `math:difference`\n- `math:greaterThan`\n- `math:lessThan`\n- `math:notLessThan`\n- `math:product`\n- `math:quotient`\n- `math:sum`\n- `string:format`\n\n## Runtime model\n\nThe generated `examples/bayes_diagnosis.js` is a specialized JavaScript derivation program. For ordinary sources, `see.js` emits the source facts as `examples/input/bayes_diagnosis.trig`. For rules-only sources, generation can reuse an existing external evidence file such as `examples/input/bayes-diagnosis.trig` or `examples/input/bayes_diagnosis.trig`. The runner reads that TriG evidence directly and performs a local fixpoint derivation; it does not parse the program source or call an external reasoner.\n\n## Output model\n\nRunning `node examples/bayes_diagnosis.js` produces a SEE-style Markdown report with an **Insight** section, an **Explanation** section, and a **Formal TriG Output** section containing the selected derived/query facts.\n";
+const DOC_MARKDOWN = "# Bayes Diagnosis\n\nGenerated by `see.js` from a Notation3 source file.\n\nNaive Bayes diagnosis example, adapted from Eyeling's bayes-diagnosis.n3.\nThe facts and rules compute unnormalized likelihoods, normalize them into\nposterior probabilities, and emit a SEE report for the highest posterior.\nValues are illustrative only and are not medical advice.\n\n## Compilation summary\n\n- Example name: `bayes_diagnosis`\n- Input facts emitted: 85\n- Forward rules compiled: 4\n- Backward predicate rules compiled: 5\n- Fuses compiled: 4\n- Predicate count: 34\n\n## Built-ins used\n\n- `list:firstRest`\n- `list:iterate`\n- `list:map`\n- `list:memberAt`\n- `log:outputString`\n- `math:difference`\n- `math:greaterThan`\n- `math:lessThan`\n- `math:notLessThan`\n- `math:product`\n- `math:quotient`\n- `math:sum`\n- `string:format`\n\n## Runtime model\n\nThe generated `examples/bayes_diagnosis.js` is a specialized JavaScript derivation program. For ordinary sources, `see.js` emits the source facts as `examples/input/bayes_diagnosis.trig`. For rules-only sources, generation can reuse an existing external evidence file such as `examples/input/bayes-diagnosis.trig` or `examples/input/bayes_diagnosis.trig`. The runner reads that TriG evidence directly and performs a local fixpoint derivation; it does not parse the program source or call an external reasoner.\n\n## Output model\n\nRunning `node examples/bayes_diagnosis.js` produces a SEE-style Markdown report with an **Entailment** section, an **Explanation** section, and a **Formal TriG Output** section containing the selected derived/query facts.\n";
 function seeMetadata(data) { return (data && data.__see) || {}; }
 function trustedDerivation(data) { const meta = seeMetadata(data); const facts = data && Array.isArray(data.facts) ? data.facts : []; const expectedFacts = EXPECTED_INPUT_FACTS || Number(meta.InputFacts || 0); if (meta.SourceSHA256 && meta.SourceSHA256 !== "46feebe5109b413a4143b7bcb4bc77ac6c0b18a8fdd8f6a62ab06c42287035d9") throw new Error('input evidence does not match the N3 source compiled into this example'); const result = saturate(facts, RULES); const rawOutput = renderRawOutput(result.graph, QUERIES, RULES, facts); fail('Compiled N3 derivation failed', { 'input evidence metadata is present and matches compiled source': meta.SourceSHA256 === "46feebe5109b413a4143b7bcb4bc77ac6c0b18a8fdd8f6a62ab06c42287035d9", 'input evidence facts were loaded': expectedFacts > 0 ? facts.length === expectedFacts : facts.length >= 0, 'compiled rules were loaded': RULES.length === 13, 'compiled query directives were loaded': QUERIES.length === 1, 'a derivation fixpoint was reached': result.graph.facts.length >= facts.length, 'query or output facts were produced': rawOutput.length > 0 }); return { ...result, rawOutput, inputFacts: facts }; }
 function snapshotMarkdown(markdown) { return markdown.split(/\n/).map((line) => line ? line + '  \n' : '\n').join(''); }

package/see/examples/bayes_therapy.js

@@ -1040,19 +1040,19 @@ function renderStructuredOutput({ title, graph, queries = [], rules = [], initia
   const lines = [];
   lines.push('# ' + title);
   lines.push('');
-  lines.push('## Insight');
+  lines.push('## Entailment');
   if (mode === 'query') {
     lines.push('The compiled query selected ' + selected.length + ' fact(s) after the rule closure was computed.');
   } else if (mode === 'formula') {
-    lines.push('The derivation produced ' + selected.length + ' formula-valued conclusion(s).');
+    lines.push('The derivation produced ' + selected.length + ' formula-valued entailment(s).');
   } else {
     lines.push('The derivation produced ' + derived.length + ' new fact(s) from ' + initialFacts.length + ' stated fact(s).');
   }
-  if (keyFact) lines.push('Main conclusion: **' + factSentence(keyFact) + '**');
+  if (keyFact) lines.push('Main entailment: **' + factSentence(keyFact) + '**');
   const bullets = selected.slice(-6).reverse();
   if (bullets.length) {
     lines.push('');
-    lines.push('Selected conclusions:');
+    lines.push('Selected entailments:');
     for (const fact of bullets) lines.push('- ' + codeFact(fact));
   }
   lines.push('');
@@ -1114,15 +1114,15 @@ function dedupeExplanationHeadings(text) {
 function normalizePublicReport(markdown, title) {
   let text = String(markdown || '').trimEnd();
   if (!/^\s*#\s+/m.test(text)) text = '# ' + title + '\n\n' + text;
-  if (!/^##\s+Insight\s*$/mi.test(text)) {
-    text = text.replace(/^(#\s+[^\n]+\n*)/, '$1\n## Insight\n');
+  if (!/^##\s+Entailment\s*$/mi.test(text)) {
+    text = text.replace(/^(#\s+[^\n]+\n*)/, '$1\n## Entailment\n');
   }
   if (!/^##\s+Explanation\s*$/mi.test(text)) {
     text += '\n\n## Explanation\nNo additional explanation was provided by the generated output.';
   }
   text = text.replace(/^##\s+([^\n]+?)\s*$/gm, (line, heading) => {
     const normalized = heading.trim().toLowerCase();
-    if (normalized === 'insight' || normalized === 'explanation') return '## ' + (normalized === 'insight' ? 'Insight' : 'Explanation');
+    if (normalized === 'insight' || normalized === 'conclusion' || normalized === 'entailment' || normalized === 'explanation') return '## ' + (normalized === 'explanation' ? 'Explanation' : 'Entailment');
     return '**' + heading.trim() + '**';
   });
   text = dedupeExplanationHeadings(text);
@@ -1131,13 +1131,13 @@ function normalizePublicReport(markdown, title) {
 function markdownize(raw, title) {
   let text = String(raw || '');
   text = text
-    .replace(/===\s*Answer\s*===/g, '## Insight')
+    .replace(/===\s*Answer\s*===/g, '## Entailment')
     .replace(/===\s*Reason\s+Why\s*===/gi, '## Explanation')
     .replace(/===\s*Explanation\s*===/gi, '## Explanation')
     .replace(/===\s*([^=]+?)\s*===/g, (_, h) => '**' + h.trim() + '**');
   text = text.replace(/^C(\d+)\s+OK\s*-\s*/gm, 'C$1: ');
   text = dedupeExplanationHeadings(text);
-  if (!text.trim()) text = '## Insight\nNo log:outputString facts were derived.\n\n## Explanation\nThe compiled derivation did not produce authored report text.';
+  if (!text.trim()) text = '## Entailment\nNo log:outputString facts were derived.\n\n## Explanation\nThe compiled derivation did not produce authored report text.';
   return normalizePublicReport(text, title);
 }
 function authoredSupportAppendix(graph, queries, rules, initialFacts, trace) {
@@ -4027,7 +4027,7 @@ const QUERIES = [
     ]
   }
 ];
-const DOC_MARKDOWN = "# Bayes Therapy\n\nGenerated by `see.js` from a Notation3 source file.\n\nExtends the Bayesian diagnosis model with a therapy utility layer. Adapted\nfrom Eyeling's bayes-therapy.n3. Values are illustrative only and are not\nmedical advice.\n\n## Compilation summary\n\n- Example name: `bayes_therapy`\n- Input facts emitted: 104\n- Forward rules compiled: 5\n- Backward predicate rules compiled: 15\n- Fuses compiled: 8\n- Predicate count: 45\n\n## Built-ins used\n\n- `list:firstRest`\n- `list:iterate`\n- `list:map`\n- `list:memberAt`\n- `log:outputString`\n- `math:difference`\n- `math:greaterThan`\n- `math:lessThan`\n- `math:notLessThan`\n- `math:product`\n- `math:quotient`\n- `math:sum`\n- `string:format`\n\n## Runtime model\n\nThe generated `examples/bayes_therapy.js` is a specialized JavaScript derivation program. For ordinary sources, `see.js` emits the source facts as `examples/input/bayes_therapy.trig`. For rules-only sources, generation can reuse an existing external evidence file such as `examples/input/bayes-therapy.trig` or `examples/input/bayes_therapy.trig`. The runner reads that TriG evidence directly and performs a local fixpoint derivation; it does not parse the program source or call an external reasoner.\n\n## Output model\n\nRunning `node examples/bayes_therapy.js` produces a SEE-style Markdown report with an **Insight** section, an **Explanation** section, and a **Formal TriG Output** section containing the selected derived/query facts.\n";
+const DOC_MARKDOWN = "# Bayes Therapy\n\nGenerated by `see.js` from a Notation3 source file.\n\nExtends the Bayesian diagnosis model with a therapy utility layer. Adapted\nfrom Eyeling's bayes-therapy.n3. Values are illustrative only and are not\nmedical advice.\n\n## Compilation summary\n\n- Example name: `bayes_therapy`\n- Input facts emitted: 104\n- Forward rules compiled: 5\n- Backward predicate rules compiled: 15\n- Fuses compiled: 8\n- Predicate count: 45\n\n## Built-ins used\n\n- `list:firstRest`\n- `list:iterate`\n- `list:map`\n- `list:memberAt`\n- `log:outputString`\n- `math:difference`\n- `math:greaterThan`\n- `math:lessThan`\n- `math:notLessThan`\n- `math:product`\n- `math:quotient`\n- `math:sum`\n- `string:format`\n\n## Runtime model\n\nThe generated `examples/bayes_therapy.js` is a specialized JavaScript derivation program. For ordinary sources, `see.js` emits the source facts as `examples/input/bayes_therapy.trig`. For rules-only sources, generation can reuse an existing external evidence file such as `examples/input/bayes-therapy.trig` or `examples/input/bayes_therapy.trig`. The runner reads that TriG evidence directly and performs a local fixpoint derivation; it does not parse the program source or call an external reasoner.\n\n## Output model\n\nRunning `node examples/bayes_therapy.js` produces a SEE-style Markdown report with an **Entailment** section, an **Explanation** section, and a **Formal TriG Output** section containing the selected derived/query facts.\n";
 function seeMetadata(data) { return (data && data.__see) || {}; }
 function trustedDerivation(data) { const meta = seeMetadata(data); const facts = data && Array.isArray(data.facts) ? data.facts : []; const expectedFacts = EXPECTED_INPUT_FACTS || Number(meta.InputFacts || 0); if (meta.SourceSHA256 && meta.SourceSHA256 !== "18244b81f37328f7d77c8e88a4749891c3d773ffbf19affecd7f118c6208395d") throw new Error('input evidence does not match the N3 source compiled into this example'); const result = saturate(facts, RULES); const rawOutput = renderRawOutput(result.graph, QUERIES, RULES, facts); fail('Compiled N3 derivation failed', { 'input evidence metadata is present and matches compiled source': meta.SourceSHA256 === "18244b81f37328f7d77c8e88a4749891c3d773ffbf19affecd7f118c6208395d", 'input evidence facts were loaded': expectedFacts > 0 ? facts.length === expectedFacts : facts.length >= 0, 'compiled rules were loaded': RULES.length === 28, 'compiled query directives were loaded': QUERIES.length === 1, 'a derivation fixpoint was reached': result.graph.facts.length >= facts.length, 'query or output facts were produced': rawOutput.length > 0 }); return { ...result, rawOutput, inputFacts: facts }; }
 function snapshotMarkdown(markdown) { return markdown.split(/\n/).map((line) => line ? line + '  \n' : '\n').join(''); }