n8n Skills

n8n Skills (by Romuald Czlonkowski, MIT-licensed) is a deep skill collection for developing, optimizing, and debugging n8n agents in production workflows. The flagship n8n-agents skill covers the n8n AI Agent node (a multi-turn LLM driver with sub-nodes for model, memory, tools, and output parser), LangChain classifier/extractor/memory patterns, RAG, tool-calling, structured output, and human-review design. It is safety-conscious by design: before activating workflows that send messages, write data, make purchases, or call external services, it instructs the agent to show exact effects and obtain approval, and to keep provider keys in n8n credentials — never in prompts or workflow JSON. A companion n8n-workflow-patterns skill covers where agents fit in a workflow at a higher level.

The honest picture: this is a community single-maintainer repo — check recency and n8n version compatibility before adopting, since n8n's LangChain node schemas change. It is n8n-specific; if you orchestrate agents elsewhere (LangGraph, raw APIs), look elsewhere. The skill assumes you already run an n8n instance.

Clone github.com/czlonkowski/n8n-skills and reference the SKILL.md files from your agent, or install via a skills aggregator.