Apple Tools MCP

/** * Real Data Integration Tests for Indexing * * These tests validate the PRODUCTION index at ~/.apple-tools-mcp/vector-index/ * Production index must be built separately via npm run rebuild-index */ import { describe, it, expect, beforeAll } from 'vitest' import fs from 'fs' import { checkDataSources, isProductionIndexReady, getProductionIndexStats, searchProductionIndex, embed, PRODUCTION_INDEX_DIR } from '../helpers/real-data.js' // Check what data sources are available const sources = checkDataSources() describe('Real Data Indexing', () => { // Skip entire suite if no data sources available or index missing describe.skipIf(!sources.mail && !sources.messages && !sources.calendar || !sources.productionIndex)( 'Full Indexing Workflow', () => { beforeAll(async () => { // Production index must already exist const ready = await isProductionIndexReady() if (!ready) { throw new Error( 'Production index not found or empty. ' + 'Run "npm run rebuild-index" to build it first.' ) } }, 60000) it('should have valid production index data', async () => { const stats = await getProductionIndexStats() // At least one table should have rows const totalRows = stats.emails + stats.messages + stats.calendar expect(totalRows).toBeGreaterThan(0) console.log('Index stats:', stats) }, 30000) it('should create production index directory', () => { expect(fs.existsSync(PRODUCTION_INDEX_DIR)).toBe(true) }) it('should have valid production index stats', async () => { const stats = await getProductionIndexStats() // At least one table should have rows const totalRows = stats.emails + stats.messages + stats.calendar expect(totalRows).toBeGreaterThan(0) console.log('Index stats:', stats) }) it('should report index as ready', async () => { const ready = await isProductionIndexReady() expect(ready).toBe(true) }) } ) }) describe.skipIf(!sources.mail)('Email Indexing (Real Data)', () => { beforeAll(async () => { // Production index must already exist const ready = await isProductionIndexReady() if (!ready) { throw new Error('Production index not found. Run "npm run rebuild-index" first.') } }, 30000) it('should index real emails', async () => { const stats = await getProductionIndexStats() expect(stats.emails).toBeGreaterThan(0) }) it('should search emails with real embeddings', async () => { // Search for a common term const results = await searchProductionIndex('meeting', 'emails', 5) // Results should have expected structure if (results.length > 0) { expect(results[0]).toHaveProperty('filePath') expect(results[0]).toHaveProperty('subject') expect(results[0]).toHaveProperty('vector') } }, 30000) it('should have correct email field structure', async () => { const stats = await getProductionIndexStats() if (stats.emails === 0) return const results = await searchProductionIndex('email', 'emails', 1) if (results.length > 0) { const email = results[0] // Check required fields expect(email).toHaveProperty('filePath') expect(email).toHaveProperty('from') expect(email).toHaveProperty('subject') expect(email).toHaveProperty('dateTimestamp') expect(email).toHaveProperty('vector') // Vector should be 384-dim expect(email.vector).toHaveLength(384) } }, 30000) }) describe.skipIf(!sources.messages)('Message Indexing (Real Data)', () => { beforeAll(async () => { const ready = await isProductionIndexReady() if (!ready) { throw new Error('Production index not found. Run "npm run rebuild-index" first.') } }, 30000) it('should index real messages', async () => { const stats = await getProductionIndexStats() expect(stats.messages).toBeGreaterThan(0) }) it('should search messages with real embeddings', async () => { const results = await searchProductionIndex('hello', 'messages', 5) if (results.length > 0) { expect(results[0]).toHaveProperty('id') expect(results[0]).toHaveProperty('text') expect(results[0]).toHaveProperty('vector') } }, 30000) it('should have correct message field structure', async () => { const stats = await getProductionIndexStats() if (stats.messages === 0) return const results = await searchProductionIndex('message', 'messages', 1) if (results.length > 0) { const message = results[0] // Check required fields expect(message).toHaveProperty('id') expect(message).toHaveProperty('text') expect(message).toHaveProperty('dateTimestamp') expect(message).toHaveProperty('vector') // Vector should be 384-dim expect(message.vector).toHaveLength(384) } }, 30000) }) describe.skipIf(!sources.calendar)('Calendar Indexing (Real Data)', () => { beforeAll(async () => { const ready = await isProductionIndexReady() if (!ready) { throw new Error('Production index not found. Run "npm run rebuild-index" first.') } }, 30000) it('should index real calendar events', async () => { const stats = await getProductionIndexStats() expect(stats.calendar).toBeGreaterThan(0) }) it('should search calendar with real embeddings', async () => { const results = await searchProductionIndex('meeting', 'calendar', 5) if (results.length > 0) { expect(results[0]).toHaveProperty('id') expect(results[0]).toHaveProperty('title') expect(results[0]).toHaveProperty('vector') } }, 30000) it('should have correct calendar field structure', async () => { const stats = await getProductionIndexStats() if (stats.calendar === 0) return const results = await searchProductionIndex('event', 'calendar', 1) if (results.length > 0) { const event = results[0] // Check required fields expect(event).toHaveProperty('id') expect(event).toHaveProperty('title') expect(event).toHaveProperty('startTimestamp') expect(event).toHaveProperty('vector') // Vector should be 384-dim expect(event.vector).toHaveLength(384) } }, 30000) }) describe('Real Embedding Quality', () => { it('should generate consistent 384-dim embeddings', async () => { const text = 'This is a test email about a meeting' const vector = await embed(text) expect(vector).toHaveLength(384) expect(vector[0]).toBeTypeOf('number') expect(Number.isFinite(vector[0])).toBe(true) }, 30000) it('should produce different embeddings for different text', async () => { const text1 = 'Meeting about quarterly budget review' const text2 = 'Dinner reservation at Italian restaurant' const vec1 = await embed(text1) const vec2 = await embed(text2) // Calculate cosine similarity let dotProduct = 0 let norm1 = 0 let norm2 = 0 for (let i = 0; i < vec1.length; i++) { dotProduct += vec1[i] * vec2[i] norm1 += vec1[i] * vec1[i] norm2 += vec2[i] * vec2[i] } const similarity = dotProduct / (Math.sqrt(norm1) * Math.sqrt(norm2)) // Unrelated text should have lower similarity (< 0.7) expect(similarity).toBeLessThan(0.7) }, 30000) it('should produce similar embeddings for similar text', async () => { const text1 = 'Schedule a meeting with John tomorrow at 2pm' const text2 = 'Set up a meeting with John tomorrow afternoon' const vec1 = await embed(text1) const vec2 = await embed(text2) // Calculate cosine similarity let dotProduct = 0 let norm1 = 0 let norm2 = 0 for (let i = 0; i < vec1.length; i++) { dotProduct += vec1[i] * vec2[i] norm1 += vec1[i] * vec1[i] norm2 += vec2[i] * vec2[i] } const similarity = dotProduct / (Math.sqrt(norm1) * Math.sqrt(norm2)) // Similar text should have high similarity (> 0.7) expect(similarity).toBeGreaterThan(0.7) }, 30000) })