AI Conversational Agent with Custom JS Tool for Random Colors

This n8n workflow demonstrates building a conversational AI agent with a custom JavaScript tool. It supports both manual testing and real-time chat triggers, processing user inputs like 'Return a random color but not green or blue' via an AI Agent node powered by OpenAI's GPT-4o-mini model. The custom Code Tool, written in JavaScript, handles the logic to select and return a random color while excluding forbidden ones, showcasing how to extend AI capabilities with code.

Key benefits include rapid prototyping of agentic AI workflows, seamless integration of LangChain nodes for chat handling and tool execution, and easy customization of tools without external dependencies. It's ideal for developers learning n8n's AI features, requiring version 1.19.4+. The workflow highlights modularity: separate triggers for testing/production, debug inputs for validation, and configurable agent prompts.

Use cases span interactive demos, customer support bots with utility tools (e.g., random generators, calculators), educational templates for AI tooling, or bases for complex agents in e-commerce (product recommenders excluding preferences) and gaming (random event generators). Deploy it to create engaging, responsive chat interfaces that leverage custom logic for dynamic responses, saving hours on boilerplate AI setup.