@gork-labs/secondbrain-mcp/dist/tools/integration-example.js

Second Brain MCP Server - Agent team orchestration with dynamic tool discovery

/**
 * Simple Integration Example for OpenAI Function Calling
 *
 * This shows how to integrate the OpenAI function calling executor into
 * the existing SecondBrain MCP server to replace regex-based tool parsing.
 *
 * Created: 2025-07-25T17:17:13+02:00
 * Author: Staff Software Engineer - Gorka
 */
import { OpenAIFunctionCallingExecutor, FunctionCallingConfig } from './openai-function-calling.js';
/**
 * Example integration in server.ts executeWithOpenAI method
 */
export class SimplifiedServerIntegration {
    mcpClientManager;
    openaiExecutor;
    constructor(mcpClientManager) {
        this.mcpClientManager = mcpClientManager;
        // Initialize OpenAI executor with API key from environment
        const apiKey = FunctionCallingConfig.getOpenAIApiKey();
        this.openaiExecutor = new OpenAIFunctionCallingExecutor(apiKey, this.mcpClientManager.getSafeTools(), 20 // maxIterations
        );
    }
    /**
     * Enhanced executeWithOpenAI method using OpenAI function calling
     */
    async executeWithOpenAI(instructions, chatmodeDefinition, sessionId) {
        // Check if this chatmode should use OpenAI function calling
        if (FunctionCallingConfig.shouldUseOpenAIFunctionCalling(chatmodeDefinition.name)) {
            // Use OpenAI native function calling
            return await this.openaiExecutor.executeSubAgent(instructions, chatmodeDefinition.name, sessionId, (toolName, args) => this.mcpClientManager.callTool(toolName, args));
        }
        else {
            // Fallback to existing regex-based implementation
            return await this.executeWithOpenAILegacy(instructions, chatmodeDefinition, sessionId);
        }
    }
    /**
     * Legacy implementation (existing regex-based approach)
     * Keep this as fallback during transition
     */
    async executeWithOpenAILegacy(instructions, chatmodeDefinition, sessionId) {
        // This would be the existing implementation
        // from server.ts lines 1850-2100
        throw new Error('Legacy implementation placeholder');
    }
}
/**
 * Environment Configuration Guide
 *
 * To use OpenAI function calling, set this environment variable:
 *
 * # Required: OpenAI API key
 * OPENAI_API_KEY=your-openai-api-key
 *
 * # Optional: Specify which chatmodes to use it for (if not set, all chatmodes)
 * SECONDBRAIN_FUNCTION_CALLING_CHATMODES=Software Engineer,Security Engineer,Database Architect
 */
/**
 * Quick Setup Steps:
 *
 * 1. Set OPENAI_API_KEY in your .env file
 * 2. Import OpenAIFunctionCallingExecutor and FunctionCallingConfig in server.ts
 * 3. Replace the executeWithOpenAI method logic with the example above
 * 4. Test with one chatmode first (e.g., Software Engineer)
 * 5. Gradually expand to other chatmodes
 *
 * Benefits:
 * - No more regex parsing errors
 * - Built-in JSON validation
 * - Better error handling
 * - Automatic retry logic
 * - Performance monitoring
 * - No feature flags needed - just works with OpenAI SDK
 */