MCP AI Bridge

import { Server } from '@modelcontextprotocol/sdk/server/index.js'; import { StdioServerTransport } from '@modelcontextprotocol/sdk/server/stdio.js'; import { CallToolRequestSchema, ListToolsRequestSchema } from '@modelcontextprotocol/sdk/types.js'; import OpenAI from 'openai'; import { GoogleGenerativeAI } from '@google/generative-ai'; import dotenv from 'dotenv'; import { homedir } from 'os'; import { join } from 'path'; import { existsSync } from 'fs'; import logger from './logger.js'; import { DEFAULTS, ERROR_MESSAGES, MODELS } from './constants.js'; import { AIBridgeError, ValidationError, ConfigurationError, APIError } from './errors.js'; import { validatePrompt, validateTemperature, validateModel, validateAPIKey } from './validators.js'; import { RateLimiter } from './rateLimiter.js'; // Try to load .env from multiple locations in order of priority const homeEnvPath = join(homedir(), '.env'); const localEnvPath = '.env'; if (existsSync(homeEnvPath)) { dotenv.config({ path: homeEnvPath }); if (process.env.NODE_ENV !== 'test') logger.info('Loaded environment from home directory'); } else if (existsSync(localEnvPath)) { dotenv.config({ path: localEnvPath }); if (process.env.NODE_ENV !== 'test') logger.info('Loaded environment from local directory'); } else { dotenv.config(); if (process.env.NODE_ENV !== 'test') logger.info('Using system environment variables'); } class AIBridgeServer { constructor() { this.server = new Server( { name: process.env.MCP_SERVER_NAME || 'AI Bridge', version: process.env.MCP_SERVER_VERSION || '1.0.0', }, { capabilities: { tools: {}, }, } ); // Validate and initialize API clients this.initializeClients(); // Initialize rate limiter this.rateLimiter = new RateLimiter(); this.setupHandlers(); this.tools = this.getAvailableTools(); } initializeClients() { // Initialize OpenAI client if (process.env.OPENAI_API_KEY) { try { validateAPIKey(process.env.OPENAI_API_KEY, 'OPENAI'); this.openai = new OpenAI({ apiKey: process.env.OPENAI_API_KEY }); if (process.env.NODE_ENV !== 'test') logger.info('OpenAI client initialized'); } catch (error) { if (process.env.NODE_ENV !== 'test') logger.error('Failed to initialize OpenAI client:', error.message); this.openai = null; } } else { this.openai = null; if (process.env.NODE_ENV !== 'test') logger.warn('OpenAI API key not provided'); } // Initialize Gemini client if (process.env.GOOGLE_AI_API_KEY) { try { this.gemini = new GoogleGenerativeAI(process.env.GOOGLE_AI_API_KEY); if (process.env.NODE_ENV !== 'test') logger.info('Gemini client initialized'); } catch (error) { if (process.env.NODE_ENV !== 'test') logger.error('Failed to initialize Gemini client:', error.message); this.gemini = null; } } else { this.gemini = null; if (process.env.NODE_ENV !== 'test') logger.warn('Google AI API key not provided'); } } getAvailableTools() { const tools = []; if (this.openai) { tools.push({ name: 'ask_openai', description: 'Ask OpenAI GPT models a question', inputSchema: { type: 'object', properties: { prompt: { type: 'string', description: 'The prompt to send to OpenAI', }, model: { type: 'string', description: `The model to use (default: ${DEFAULTS.OPENAI.MODEL})`, enum: MODELS.OPENAI, default: DEFAULTS.OPENAI.MODEL, }, temperature: { type: 'number', description: `Temperature for response generation (${DEFAULTS.OPENAI.MIN_TEMPERATURE}-${DEFAULTS.OPENAI.MAX_TEMPERATURE})`, default: DEFAULTS.OPENAI.TEMPERATURE, minimum: DEFAULTS.OPENAI.MIN_TEMPERATURE, maximum: DEFAULTS.OPENAI.MAX_TEMPERATURE, }, }, required: ['prompt'], }, }); } if (this.gemini) { tools.push({ name: 'ask_gemini', description: 'Ask Google Gemini AI a question', inputSchema: { type: 'object', properties: { prompt: { type: 'string', description: 'The prompt to send to Gemini', }, model: { type: 'string', description: `The model to use (default: ${DEFAULTS.GEMINI.MODEL})`, enum: MODELS.GEMINI, default: DEFAULTS.GEMINI.MODEL, }, temperature: { type: 'number', description: `Temperature for response generation (${DEFAULTS.GEMINI.MIN_TEMPERATURE}-${DEFAULTS.GEMINI.MAX_TEMPERATURE})`, default: DEFAULTS.GEMINI.TEMPERATURE, minimum: DEFAULTS.GEMINI.MIN_TEMPERATURE, maximum: DEFAULTS.GEMINI.MAX_TEMPERATURE, }, }, required: ['prompt'], }, }); } tools.push({ name: 'server_info', description: 'Get server status and configuration', inputSchema: { type: 'object', properties: {}, }, }); return tools; } setupHandlers() { this.server.setRequestHandler(ListToolsRequestSchema, async () => ({ tools: this.tools, })); this.server.setRequestHandler(CallToolRequestSchema, async (request) => { const { name, arguments: args } = request.params; try { // Check rate limit this.rateLimiter.checkLimit(); switch (name) { case 'ask_openai': return await this.handleOpenAI(args); case 'ask_gemini': return await this.handleGemini(args); case 'server_info': return this.handleServerInfo(); default: throw new ValidationError(`${ERROR_MESSAGES.UNKNOWN_TOOL}: ${name}`); } } catch (error) { if (process.env.NODE_ENV !== 'test') logger.error(`Error handling ${name}:`, error); return this.formatError(error); } }); } async handleOpenAI(args) { if (!this.openai) { throw new ConfigurationError(ERROR_MESSAGES.OPENAI_NOT_CONFIGURED); } // Validate inputs const prompt = validatePrompt(args.prompt); const model = validateModel(args.model, 'OPENAI'); const temperature = validateTemperature(args.temperature, 'OPENAI'); try { if (process.env.NODE_ENV !== 'test') logger.debug(`OpenAI request - model: ${model}, temperature: ${temperature}`); const completion = await this.openai.chat.completions.create({ model: model, messages: [{ role: 'user', content: prompt }], temperature: temperature, }); return { content: [ { type: 'text', text: `🤖 OPENAI RESPONSE (${model}):\n\n${completion.choices[0].message.content}`, }, ], }; } catch (error) { if (error.status === 429) { throw new APIError('OpenAI rate limit exceeded. Please try again later.', 'OpenAI'); } else if (error.status === 401) { throw new ConfigurationError('Invalid OpenAI API key'); } else { throw new APIError(`OpenAI API error: ${error.message}`, 'OpenAI'); } } } async handleGemini(args) { if (!this.gemini) { throw new ConfigurationError(ERROR_MESSAGES.GEMINI_NOT_CONFIGURED); } // Validate inputs const prompt = validatePrompt(args.prompt); const model = validateModel(args.model, 'GEMINI'); const temperature = validateTemperature(args.temperature, 'GEMINI'); try { if (process.env.NODE_ENV !== 'test') logger.debug(`Gemini request - model: ${model}, temperature: ${temperature}`); const geminiModel = this.gemini.getGenerativeModel({ model: model, generationConfig: { temperature: temperature, }, }); const result = await geminiModel.generateContent(prompt); const response = await result.response; const text = response.text(); return { content: [ { type: 'text', text: `🤖 GEMINI RESPONSE (${model}):\n\n${text}`, }, ], }; } catch (error) { if (error.message?.includes('quota')) { throw new APIError('Gemini quota exceeded. Please try again later.', 'Gemini'); } else if (error.message?.includes('API key')) { throw new ConfigurationError('Invalid Gemini API key'); } else { throw new APIError(`Gemini API error: ${error.message}`, 'Gemini'); } } } handleServerInfo() { const info = { name: process.env.MCP_SERVER_NAME || 'AI Bridge', version: process.env.MCP_SERVER_VERSION || '1.0.0', openai: { configured: !!this.openai, models: this.openai ? MODELS.OPENAI : [], }, gemini: { configured: !!this.gemini, models: this.gemini ? MODELS.GEMINI : [], }, rateLimits: { maxRequests: DEFAULTS.RATE_LIMIT.MAX_REQUESTS, windowMs: DEFAULTS.RATE_LIMIT.WINDOW_MS, }, security: { inputValidation: true, rateLimiting: true, promptMaxLength: DEFAULTS.PROMPT.MAX_LENGTH, }, }; return { content: [ { type: 'text', text: `🤖 AI BRIDGE SERVER INFO:\n\n${JSON.stringify(info, null, 2)}`, }, ], }; } formatError(error) { // Don't expose stack traces or sensitive information const message = error instanceof AIBridgeError ? error.message : 'An unexpected error occurred'; return { content: [ { type: 'text', text: `Error: ${message}`, }, ], }; } async run() { const transport = new StdioServerTransport(); await this.server.connect(transport); if (process.env.NODE_ENV !== 'test') logger.info('AI Bridge MCP server running'); } } // Export for testing export { AIBridgeServer }; // Run the server if this is the main module // Only run in production, not during tests if (process.env.NODE_ENV !== 'test' && process.env.NODE_ENV !== 'development') { const server = new AIBridgeServer(); server.run().catch((error) => { if (process.env.NODE_ENV !== 'test') logger.error('Failed to start server:', error); process.exit(1); }); }