Custom JS LangChain Agent: OpenAI Joke & Wikipedia Facts

This workflow showcases a custom LangChain agent implemented in JavaScript within n8n, triggered manually via 'Execute Workflow'. It branches into two paths for experimentation with LLMs. The first path processes a preset 'Tell me a joke' input through a self-coded LLM Chain node, interfacing with OpenAI to generate humorous responses. The second path handles 'What year was Einstein born?' via an Agent node that leverages Chat OpenAI and a custom Wikipedia tool for accurate information retrieval.

Key features include built-in nodes like Manual Trigger and Set, alongside custom nodes for advanced LLM chaining and tool integration. Sticky notes provide documentation on self-coded components. Requires n8n 1.19.4+. Benefits: Rapid prototyping of conversational AI agents, seamless OpenAI integration, and extensible tool usage without external coding.

Ideal use cases: AI developers testing agent architectures, educators demoing LLM capabilities, or teams building info-retrieval bots. Saves hours on setup for multi-tool AI workflows, enabling focus on logic refinement and scaling to production agents.