All projects · AI Agents & RAG
Q&A Test — Copywriter Agent over Webhook
A live HTTP API endpoint: a POST webhook receives a JSON body with a prompt, validates it, and routes valid requests to a LangChain AI Agent…
A live HTTP API endpoint: a POST webhook receives a JSON body with a prompt, validates it, and routes valid requests to a LangChain AI Agent (GPT-4o-mini) configured as a direct-response B2B copywriter with session-keyed conversation memory. The agent's copy is shaped into a JSON reply and returned to the caller; missing-prompt requests get a JSON error response.
- Full request/response webhook API pattern with responseMode responseNode
- If node validates body.prompt (exists AND notEmpty) and branches to a dedicated JSON error responder
- LangChain AI Agent with a Simple Memory buffer keyed by body.sessionId for multi-turn context
- Outcome-selling copywriter system prompt (sell the business result, never 'a website/app')
- Workflow is active:true — the only live/deployed workflow of the five
Built with: n8n, Webhook / Respond to Webhook, @n8n/n8n-nodes-langchain.agent, OpenAI gpt-4o-mini (lmChatOpenAi), memoryBufferWindow (Simple Memory), If branching, JavaScript Code node