ThoughtMCP

/** * Tests for ConfidenceCommunicationModule * * Tests the confidence communication module that provides clear confidence * presentation, interpretation guidance, uncertainty explanations, action * recommendations, and factor breakdown displays. * * Following TDD: These tests are written FIRST and should FAIL initially. */ import { beforeEach, describe, expect, it } from "vitest"; import { ConfidenceCommunicationModule } from "../../../confidence/confidence-communication"; import type { ConfidenceAssessment } from "../../../confidence/types"; import { UncertaintyType as UncertaintyTypeEnum } from "../../../confidence/types"; describe("ConfidenceCommunicationModule", () => { let module: ConfidenceCommunicationModule; beforeEach(() => { module = new ConfidenceCommunicationModule(); }); describe("formatConfidence", () => { it("should format high confidence (0.8-1.0) with percentage and label", () => { const result = module.formatConfidence(0.9); expect(result).toContain("90%"); expect(result.toLowerCase()).toMatch(/high|strong|very confident/); }); it("should format medium confidence (0.5-0.8) with percentage and label", () => { const result = module.formatConfidence(0.65); expect(result).toContain("65%"); expect(result.toLowerCase()).toMatch(/medium|moderate|fairly confident/); }); it("should format fairly confident range (0.65-0.8)", () => { const result = module.formatConfidence(0.7); expect(result).toContain("70%"); expect(result).toContain("Fairly Confident"); }); it("should format very low confidence (0-0.3)", () => { const result = module.formatConfidence(0.15); expect(result).toContain("15%"); expect(result).toContain("Very Low"); }); it("should format low confidence (0-0.5) with percentage and label", () => { const result = module.formatConfidence(0.3); expect(result).toContain("30%"); expect(result.toLowerCase()).toMatch(/low|weak|uncertain/); }); it("should handle edge case: confidence = 0", () => { const result = module.formatConfidence(0); expect(result).toContain("0%"); expect(result.toLowerCase()).toMatch(/very low|no confidence|highly uncertain/); }); it("should handle edge case: confidence = 1", () => { const result = module.formatConfidence(1); expect(result).toContain("100%"); expect(result.toLowerCase()).toMatch(/very high|complete|certain/); }); it("should handle edge case: confidence = 0.5 (boundary)", () => { const result = module.formatConfidence(0.5); expect(result).toContain("50%"); expect(result).toBeTruthy(); }); it("should throw error for invalid confidence < 0", () => { expect(() => module.formatConfidence(-0.1)).toThrow(); }); it("should throw error for invalid confidence > 1", () => { expect(() => module.formatConfidence(1.1)).toThrow(); }); }); describe("provideInterpretation", () => { const createMockContext = (domain: string = "general") => ({ domain, problemType: "analysis", }); it("should provide interpretation for high confidence", () => { const interpretation = module.provideInterpretation(0.9, createMockContext()); expect(interpretation).toBeTruthy(); expect(interpretation.length).toBeGreaterThan(20); expect(interpretation.toLowerCase()).toMatch(/reliable|trust|confident|strong/); }); it("should provide interpretation for medium confidence", () => { const interpretation = module.provideInterpretation(0.6, createMockContext()); expect(interpretation).toBeTruthy(); expect(interpretation.length).toBeGreaterThan(20); expect(interpretation.toLowerCase()).toMatch(/moderate|reasonable|consider|verify/); }); it("should provide interpretation for low confidence", () => { const interpretation = module.provideInterpretation(0.3, createMockContext()); expect(interpretation).toBeTruthy(); expect(interpretation.length).toBeGreaterThan(20); expect(interpretation.toLowerCase()).toMatch(/uncertain|caution|more information|verify/); }); it("should provide interpretation for moderate confidence (0.4-0.6)", () => { const interpretation = module.provideInterpretation(0.5, createMockContext()); expect(interpretation).toBeTruthy(); expect(interpretation.length).toBeGreaterThan(20); expect(interpretation.toLowerCase()).toMatch(/moderate|uncertainty|gaps|information/); }); it("should provide context-specific interpretation for different domains", () => { const medicalContext = createMockContext("medical"); const financialContext = createMockContext("financial"); const medicalInterpretation = module.provideInterpretation(0.6, medicalContext); const financialInterpretation = module.provideInterpretation(0.6, financialContext); expect(medicalInterpretation).toBeTruthy(); expect(financialInterpretation).toBeTruthy(); // Interpretations should be meaningful for their context expect(medicalInterpretation.length).toBeGreaterThan(20); expect(financialInterpretation.length).toBeGreaterThan(20); }); it("should handle edge case: confidence = 0", () => { const interpretation = module.provideInterpretation(0, createMockContext()); expect(interpretation).toBeTruthy(); expect(interpretation.toLowerCase()).toMatch(/no confidence|highly uncertain|not reliable/); }); it("should handle edge case: confidence = 1", () => { const interpretation = module.provideInterpretation(1, createMockContext()); expect(interpretation).toBeTruthy(); expect(interpretation.toLowerCase()).toMatch(/very confident|highly reliable|certain/); }); }); describe("explainUncertainty", () => { it("should explain epistemic uncertainty clearly", () => { const explanation = module.explainUncertainty(UncertaintyTypeEnum.EPISTEMIC); expect(explanation).toBeTruthy(); expect(explanation.length).toBeGreaterThan(30); expect(explanation.toLowerCase()).toMatch(/knowledge|information|learn|reduce/); expect(explanation).not.toMatch(/technical jargon|complex terms/); }); it("should explain aleatory uncertainty clearly", () => { const explanation = module.explainUncertainty(UncertaintyTypeEnum.ALEATORY); expect(explanation).toBeTruthy(); expect(explanation.length).toBeGreaterThan(30); expect(explanation.toLowerCase()).toMatch(/random|variability|inherent|cannot be reduced/); expect(explanation).not.toMatch(/technical jargon|complex terms/); }); it("should explain ambiguity uncertainty clearly", () => { const explanation = module.explainUncertainty(UncertaintyTypeEnum.AMBIGUITY); expect(explanation).toBeTruthy(); expect(explanation.length).toBeGreaterThan(30); expect(explanation.toLowerCase()).toMatch(/multiple|interpretation|ambiguous|clarify/); expect(explanation).not.toMatch(/technical jargon|complex terms/); }); it("should provide different explanations for different uncertainty types", () => { const epistemicExplanation = module.explainUncertainty(UncertaintyTypeEnum.EPISTEMIC); const aleatoryExplanation = module.explainUncertainty(UncertaintyTypeEnum.ALEATORY); const ambiguityExplanation = module.explainUncertainty(UncertaintyTypeEnum.AMBIGUITY); expect(epistemicExplanation).not.toBe(aleatoryExplanation); expect(epistemicExplanation).not.toBe(ambiguityExplanation); expect(aleatoryExplanation).not.toBe(ambiguityExplanation); }); }); describe("recommendActions", () => { it("should recommend actions for high confidence", () => { const actions = module.recommendActions(0.9); expect(Array.isArray(actions)).toBe(true); expect(actions.length).toBeGreaterThan(0); expect(actions.length).toBeLessThanOrEqual(5); expect(actions.some((a: string) => a.toLowerCase().includes("proceed"))).toBe(true); }); it("should recommend actions for medium confidence", () => { const actions = module.recommendActions(0.6); expect(Array.isArray(actions)).toBe(true); expect(actions.length).toBeGreaterThan(0); expect(actions.length).toBeLessThanOrEqual(5); expect( actions.some((a: string) => a.toLowerCase().match(/verify|validate|review|consider/)) ).toBe(true); }); it("should recommend actions for low confidence", () => { const actions = module.recommendActions(0.3); expect(Array.isArray(actions)).toBe(true); expect(actions.length).toBeGreaterThan(0); expect(actions.length).toBeLessThanOrEqual(5); expect( actions.some((a: string) => a.toLowerCase().match(/more information|gather|investigate|caution/) ) ).toBe(true); }); it("should recommend actions for very low confidence", () => { const actions = module.recommendActions(0.1); expect(Array.isArray(actions)).toBe(true); expect(actions.length).toBeGreaterThan(0); expect( actions.some((a: string) => a.toLowerCase().match(/do not proceed|seek expert|more data/)) ).toBe(true); }); it("should recommend actions for confidence in 0.2-0.3 range", () => { const actions = module.recommendActions(0.25); expect(Array.isArray(actions)).toBe(true); expect(actions.length).toBeGreaterThan(0); expect( actions.some((a: string) => a.toLowerCase().match(/do not proceed|substantial|research|expert/) ) ).toBe(true); }); it("should provide actionable recommendations (not vague)", () => { const actions = module.recommendActions(0.5); actions.forEach((action: string) => { expect(action.length).toBeGreaterThan(10); expect(action).toMatch(/[A-Z]/); // Should start with capital letter }); }); it("should handle edge case: confidence = 0", () => { const actions = module.recommendActions(0); expect(Array.isArray(actions)).toBe(true); expect(actions.length).toBeGreaterThan(0); }); it("should handle edge case: confidence = 1", () => { const actions = module.recommendActions(1); expect(Array.isArray(actions)).toBe(true); expect(actions.length).toBeGreaterThan(0); }); }); describe("generateFactorBreakdown", () => { const createMockAssessment = (overallConfidence: number): ConfidenceAssessment => ({ overallConfidence, evidenceQuality: 0.8, reasoningCoherence: 0.7, completeness: 0.6, uncertaintyLevel: 0.3, uncertaintyType: UncertaintyTypeEnum.EPISTEMIC, factors: [ { dimension: "evidence", score: 0.8, weight: 0.3, explanation: "Strong evidence available", }, { dimension: "coherence", score: 0.7, weight: 0.3, explanation: "Reasoning is mostly coherent", }, { dimension: "completeness", score: 0.6, weight: 0.25, explanation: "Some gaps in analysis", }, { dimension: "uncertainty", score: 0.7, weight: 0.15, explanation: "Moderate uncertainty present", }, ], timestamp: new Date(), processingTime: 50, }); it("should generate factor breakdown with overall confidence", () => { const assessment = createMockAssessment(0.7); const breakdown = module.generateFactorBreakdown(assessment); expect(breakdown).toBeDefined(); expect(breakdown.overall).toBe(0.7); }); it("should include all confidence factors", () => { const assessment = createMockAssessment(0.7); const breakdown = module.generateFactorBreakdown(assessment); expect(breakdown.factors).toBeDefined(); expect(Array.isArray(breakdown.factors)).toBe(true); expect(breakdown.factors.length).toBeGreaterThan(0); expect(breakdown.factors.length).toBe(assessment.factors.length); }); it("should identify strengths (high-scoring factors)", () => { const assessment = createMockAssessment(0.7); const breakdown = module.generateFactorBreakdown(assessment); expect(breakdown.strengths).toBeDefined(); expect(Array.isArray(breakdown.strengths)).toBe(true); expect(breakdown.strengths.length).toBeGreaterThan(0); expect(breakdown.strengths.some((s: string) => s.toLowerCase().includes("evidence"))).toBe( true ); }); it("should identify weaknesses (low-scoring factors)", () => { const assessment = createMockAssessment(0.7); const breakdown = module.generateFactorBreakdown(assessment); expect(breakdown.weaknesses).toBeDefined(); expect(Array.isArray(breakdown.weaknesses)).toBe(true); expect(breakdown.weaknesses.length).toBeGreaterThan(0); expect( breakdown.weaknesses.some((w: string) => w.toLowerCase().includes("completeness")) ).toBe(true); }); it("should provide recommendations to improve confidence", () => { const assessment = createMockAssessment(0.7); const breakdown = module.generateFactorBreakdown(assessment); expect(breakdown.recommendations).toBeDefined(); expect(Array.isArray(breakdown.recommendations)).toBe(true); expect(breakdown.recommendations.length).toBeGreaterThan(0); expect(breakdown.recommendations.length).toBeLessThanOrEqual(5); }); it("should handle assessment with all high scores", () => { const assessment: ConfidenceAssessment = { overallConfidence: 0.95, evidenceQuality: 0.95, reasoningCoherence: 0.95, completeness: 0.95, uncertaintyLevel: 0.05, uncertaintyType: UncertaintyTypeEnum.EPISTEMIC, factors: [ { dimension: "evidence", score: 0.95, weight: 0.3, explanation: "Excellent evidence", }, { dimension: "coherence", score: 0.95, weight: 0.3, explanation: "Highly coherent", }, { dimension: "completeness", score: 0.95, weight: 0.25, explanation: "Very complete", }, { dimension: "uncertainty", score: 0.95, weight: 0.15, explanation: "Very low uncertainty", }, ], timestamp: new Date(), processingTime: 50, }; const breakdown = module.generateFactorBreakdown(assessment); expect(breakdown.strengths.length).toBeGreaterThan(0); expect(breakdown.weaknesses.length).toBe(0); }); it("should handle assessment with all low scores", () => { const assessment: ConfidenceAssessment = { overallConfidence: 0.2, evidenceQuality: 0.2, reasoningCoherence: 0.2, completeness: 0.2, uncertaintyLevel: 0.8, uncertaintyType: UncertaintyTypeEnum.EPISTEMIC, factors: [ { dimension: "evidence", score: 0.2, weight: 0.3, explanation: "Weak evidence", }, { dimension: "coherence", score: 0.2, weight: 0.3, explanation: "Incoherent reasoning", }, { dimension: "completeness", score: 0.2, weight: 0.25, explanation: "Incomplete analysis", }, { dimension: "uncertainty", score: 0.2, weight: 0.15, explanation: "High uncertainty", }, ], timestamp: new Date(), processingTime: 50, }; const breakdown = module.generateFactorBreakdown(assessment); expect(breakdown.weaknesses.length).toBeGreaterThan(0); expect(breakdown.strengths.length).toBe(0); }); it("should handle edge case: empty factors array", () => { const assessment: ConfidenceAssessment = { overallConfidence: 0.5, evidenceQuality: 0.5, reasoningCoherence: 0.5, completeness: 0.5, uncertaintyLevel: 0.5, uncertaintyType: UncertaintyTypeEnum.EPISTEMIC, factors: [], timestamp: new Date(), processingTime: 50, }; const breakdown = module.generateFactorBreakdown(assessment); expect(breakdown).toBeDefined(); expect(breakdown.overall).toBe(0.5); expect(breakdown.factors).toEqual([]); }); it("should recommend gathering evidence when evidenceQuality < 0.6", () => { const assessment: ConfidenceAssessment = { overallConfidence: 0.5, evidenceQuality: 0.4, reasoningCoherence: 0.7, completeness: 0.7, uncertaintyLevel: 0.3, uncertaintyType: UncertaintyTypeEnum.EPISTEMIC, factors: [], timestamp: new Date(), processingTime: 50, }; const breakdown = module.generateFactorBreakdown(assessment); expect( breakdown.recommendations.some((r: string) => r.toLowerCase().includes("evidence")) ).toBe(true); }); it("should recommend reviewing reasoning when reasoningCoherence < 0.6", () => { const assessment: ConfidenceAssessment = { overallConfidence: 0.5, evidenceQuality: 0.7, reasoningCoherence: 0.4, completeness: 0.7, uncertaintyLevel: 0.3, uncertaintyType: UncertaintyTypeEnum.EPISTEMIC, factors: [], timestamp: new Date(), processingTime: 50, }; const breakdown = module.generateFactorBreakdown(assessment); expect( breakdown.recommendations.some((r: string) => r.toLowerCase().includes("reasoning")) ).toBe(true); }); it("should recommend addressing gaps when completeness < 0.6", () => { const assessment: ConfidenceAssessment = { overallConfidence: 0.5, evidenceQuality: 0.7, reasoningCoherence: 0.7, completeness: 0.4, uncertaintyLevel: 0.3, uncertaintyType: UncertaintyTypeEnum.EPISTEMIC, factors: [], timestamp: new Date(), processingTime: 50, }; const breakdown = module.generateFactorBreakdown(assessment); expect( breakdown.recommendations.some((r: string) => r.toLowerCase().match(/gaps|completeness/)) ).toBe(true); }); it("should recommend reducing uncertainty when uncertaintyLevel > 0.5", () => { const assessment: ConfidenceAssessment = { overallConfidence: 0.5, evidenceQuality: 0.7, reasoningCoherence: 0.7, completeness: 0.7, uncertaintyLevel: 0.7, uncertaintyType: UncertaintyTypeEnum.EPISTEMIC, factors: [], timestamp: new Date(), processingTime: 50, }; const breakdown = module.generateFactorBreakdown(assessment); expect( breakdown.recommendations.some((r: string) => r.toLowerCase().includes("uncertainty")) ).toBe(true); }); }); describe("Edge Cases and Error Handling", () => { it("should handle null confidence gracefully", () => { expect(() => module.formatConfidence(null as unknown as number)).toThrow(); }); it("should handle undefined confidence gracefully", () => { expect(() => module.formatConfidence(undefined as unknown as number)).toThrow(); }); it("should handle NaN confidence gracefully", () => { expect(() => module.formatConfidence(NaN)).toThrow(); }); it("should handle Infinity confidence gracefully", () => { expect(() => module.formatConfidence(Infinity)).toThrow(); }); it("should handle negative Infinity confidence gracefully", () => { expect(() => module.formatConfidence(-Infinity)).toThrow(); }); }); });