ThoughtMCP

#!/usr/bin/env node /** * REST API Server Entry Point * * Standalone entry point for starting the REST API server. * This script initializes the cognitive core and starts the REST API server * to run alongside the MCP server in the same Docker container. * * Requirements: 16.1, 16.2, 16.3 * * Environment Variables: * REST_API_PORT - Port for REST API server (default: 3000) * CORS_ORIGINS - Comma-separated list of allowed CORS origins (default: http://localhost:5173) * DB_HOST - Database host (default: localhost) * DB_PORT - Database port (default: 5432) * DB_NAME - Database name (default: thoughtmcp) * DB_USER - Database user (default: postgres) * DB_PASSWORD - Database password * OLLAMA_HOST - Ollama API URL (default: http://localhost:11434) * EMBEDDING_MODEL - Embedding model name (default: nomic-embed-text) */ import { BiasPatternRecognizer } from "./bias/bias-pattern-recognizer.js"; import { MultiDimensionalConfidenceAssessor } from "./confidence/multi-dimensional-assessor.js"; import { DatabaseConnectionManager } from "./database/connection-manager.js"; import { GenericLRUCache } from "./embeddings/cache.js"; import { EmbeddingEngine } from "./embeddings/embedding-engine.js"; import { EmbeddingStorage } from "./embeddings/embedding-storage.js"; import { OllamaEmbeddingModel } from "./embeddings/models/ollama-model.js"; import { CircumplexEmotionAnalyzer } from "./emotion/circumplex-analyzer.js"; import type { EmotionModel } from "./emotion/types.js"; import { FrameworkRegistry } from "./framework/framework-registry.js"; import { FrameworkSelector } from "./framework/framework-selector.js"; import { ProblemClassifier } from "./framework/problem-classifier.js"; import { WaypointGraphBuilder } from "./graph/waypoint-builder.js"; import { MemoryRepository } from "./memory/memory-repository.js"; import { MemoryAugmentedReasoning } from "./reasoning/memory-augmented-reasoning.js"; import { ParallelReasoningOrchestrator } from "./reasoning/orchestrator.js"; import { ProblemDecomposer } from "./reasoning/problem-decomposer.js"; import { type CognitiveCore, initializeSharedCore, isSharedCoreInitialized, } from "./server/cognitive-core.js"; import { RestApiServer, type RestApiServerConfig } from "./server/rest-api-server.js"; import { Logger } from "./utils/logger.js"; /** * Parse CORS origins from environment variable * Supports comma-separated list of origins */ function parseCorsOrigins(originsEnv: string | undefined): string[] { if (!originsEnv) { return ["http://localhost:5173"]; } return originsEnv.split(",").map((origin) => origin.trim()); } /** * Initialize the cognitive core with all components * Requirements: 16.2, 16.4 */ async function initializeCognitiveCore(): Promise<CognitiveCore> { Logger.info("Initializing cognitive core for REST API..."); // Initialize database connection const databaseManager = new DatabaseConnectionManager({ host: process.env.DB_HOST ?? "localhost", port: parseInt(process.env.DB_PORT ?? "5432"), database: process.env.DB_NAME ?? "thoughtmcp", user: process.env.DB_USER ?? "postgres", password: process.env.DB_PASSWORD ?? "", poolSize: parseInt(process.env.DB_POOL_SIZE ?? "20"), connectionTimeout: 5000, idleTimeout: 30000, }); await databaseManager.connect(); Logger.info("Database connection established"); // Initialize embedding engine const embeddingModel = process.env.EMBEDDING_MODEL ?? "nomic-embed-text"; const embeddingDimension = parseInt(process.env.EMBEDDING_DIMENSION ?? "768"); const model = new OllamaEmbeddingModel({ host: process.env.OLLAMA_HOST ?? "http://localhost:11434", modelName: embeddingModel, dimension: embeddingDimension, timeout: 30000, maxRetries: 3, }); const cache = new GenericLRUCache<number[]>(10000, 3600000); const embeddingEngine = new EmbeddingEngine(model, cache); Logger.info("Embedding engine initialized"); // Initialize embedding storage and graph builder const embeddingStorage = new EmbeddingStorage(databaseManager); const graphBuilder = new WaypointGraphBuilder(databaseManager, embeddingStorage, { similarityThreshold: 0.7, maxLinksPerNode: 3, minLinksPerNode: 1, enableBidirectional: true, }); // Initialize memory repository const memoryRepository = new MemoryRepository( databaseManager, embeddingEngine, graphBuilder, embeddingStorage ); Logger.info("Memory repository initialized"); // Initialize reasoning components const reasoningOrchestrator = new ParallelReasoningOrchestrator(); const problemDecomposer = new ProblemDecomposer(); const memoryAugmentedReasoning = new MemoryAugmentedReasoning(memoryRepository, { minSalience: 0.5, maxMemories: 10, minStrength: 0.3, }); Logger.info("Reasoning components initialized"); // Initialize framework selector const classifier = new ProblemClassifier(); const registry = FrameworkRegistry.getInstance(); const frameworkSelector = new FrameworkSelector(classifier, registry); Logger.info("Framework selector initialized"); // Initialize metacognitive components const confidenceAssessor = new MultiDimensionalConfidenceAssessor(); const biasDetector = new BiasPatternRecognizer(); Logger.info("Metacognitive components initialized"); // Initialize emotion analyzer const emotionModel: EmotionModel = { name: "circumplex-v1", version: "1.0.0", }; const emotionAnalyzer = new CircumplexEmotionAnalyzer(emotionModel); Logger.info("Emotion analyzer initialized"); // Create cognitive core const cognitiveCore: CognitiveCore = { memoryRepository, reasoningOrchestrator, frameworkSelector, confidenceAssessor, biasDetector, emotionAnalyzer, problemDecomposer, memoryAugmentedReasoning, }; // Initialize shared core singleton if (!isSharedCoreInitialized()) { initializeSharedCore(cognitiveCore); } Logger.info("Cognitive core initialized successfully"); return cognitiveCore; } /** * Main entry point for REST API server * Requirements: 16.1, 16.3 */ async function main(): Promise<void> { Logger.info("Starting REST API server..."); try { // Initialize cognitive core first (Requirements: 16.2) const cognitiveCore = await initializeCognitiveCore(); // Parse configuration from environment const port = parseInt(process.env.REST_API_PORT ?? "3000"); const corsOrigins = parseCorsOrigins(process.env.CORS_ORIGINS); const config: Partial<RestApiServerConfig> = { port, corsOrigins, rateLimitWindow: parseInt(process.env.RATE_LIMIT_WINDOW ?? "60000"), rateLimitMax: parseInt(process.env.RATE_LIMIT_MAX ?? "100"), enableWebSocket: process.env.ENABLE_WEBSOCKET !== "false", enableSSE: process.env.ENABLE_SSE !== "false", }; Logger.info(`REST API configuration: port=${port}, corsOrigins=${corsOrigins.join(",")}`); // Create and start REST API server const server = new RestApiServer(cognitiveCore, config); // Mount routes and error handlers server.mountRoutes(); server.mountErrorHandlers(); // Setup graceful shutdown server.setupGracefulShutdown(); // Start the server await server.start(); Logger.info(`REST API server started successfully on port ${port}`); Logger.info("REST API is ready to accept connections"); } catch (error) { Logger.error("Failed to start REST API server:", error); process.exit(1); } } // Run main if executed directly const isMainModule = import.meta.url === `file://${process.argv[1]}` || (process.argv[1]?.endsWith("rest-api-start.js") ?? false); if (isMainModule) { main().catch((error) => { Logger.error("Fatal error:", error); process.exit(1); }); } export { initializeCognitiveCore, main, parseCorsOrigins };