n8n-agents
Design n8n AI agents, chains, classifiers, extractors, tool calling, memory, RAG, structured output, and human-review flows.
Profile →npx skills add sickn33/agentic-awesome-skills --skill n8n-agents --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
Designs n8n AI agents, chains, classifiers, extractors, tool calling, memory, RAG, structured output, and human-review flows.
How it works
The skill provides guidance for building an n8n AI Agent workflow using the node @n8n/n8n-nodes-langchain.agent with sub-nodes for model, memory, tools, and an optional output parser. It instructs choosing appropriate sub-nodes (e.g., .agent, .lmChatOpenAi, .memoryBufferWindow, .toolWorkflow, .outputParserStructured) and wiring them via ai_* connections in workflow JSON. It details how to wire tools, memory, and output parsers, and how to structure the agent's inputs and outputs (e.g., placing the final answer in $json.output). It emphasizes the correct use of tool descriptions, the role of structured output with autoFix, and the need for human review for certain tool actions. It also covers boundaries between the agent and binary data, and how to stage media externally when needed. Finally, it explains the four sub-node slots (model, memory, tools, outputParser) and how they connect from sub-nodes to the agent.
When to use it
Use this skill for designing n8n AI Agents, LangChain integrations, classifiers, extractors, memory/RAG setups, tool-calling, structured output, or human-review flows. It cautions to confirm the target n8n instance and inspect the live node schema before applying version-sensitive configuration, and to obtain explicit user approval for potentially destructive actions before execution. It also advises on when to use specific node types and how to route inputs to appropriate branches.
What it can touch
The skill references and governs use of tools within the n8n LangChain ecosystem, including tool nodes like .toolWorkflow, .toolHttpRequest, .toolCode, and memory nodes like .memoryBufferWindow, .memoryPostgresChat, as well as an outputParserStructured node. It specifies that tool parameters may use $fromAI() to populate inputs, and that the final agent output is exposed as $json.output for downstream nodes. It also discusses the need to avoid binary payloads in tool calls and to stage media externally when necessary.
Caveats
It notes the importance of not placing provider keys in prompts or workflow JSON, and to gate sensitive actions behind human review when appropriate. It emphasizes structured output with autoFix: true and using a real JSON Schema for schemas. It warns about potential infinite loops in chat agents if replies aren’t filtered, and about the boundary between the agent and binary data handling. These points are presented as design guidance and operational cautions rather than guarantees of success.
# n8n Agents ## When to Use Use this skill for n8n AI Agent, LangChain, classifier, extractor, memory, RAG, tool-calling, structured-output, or human-review design. Confirm the target n8n instance and inspect the live node schema before applying version-sensitive configuration. Before activating or testing a workflow that can send messages, write data, make purchases, change accounts, or call external services, show the user the exact effects and obtain approval. Store provider keys and tokens only in n8n credentials; never place them in prompts, Set nodes, workflow JSON, examples, or logs. The n8n AI Agent node (`@n8n/n8n-nodes-langchain.agent`) is a multi-turn LLM driver with sub-nodes for the model, memory, tools, and an optional output parser. This skill is the **deep** guide to designing agents and the LangChain family around them. For the high-level "where an agent fits in a workflow" picture, see the **n8n-workflow-patterns** skill — this skill goes one level down into *how to build it well*. For node-type formats: in workflow JSON the LangChain nodes use the long `@n8n/n8n-nodes-langchain.*` form (`.agent`, `.lmChatOpenAi`, `.memoryBufferWindow`, `.outputParserStructured`,
- When to Use
- Pick the right node first
- The sub-node pattern
- Two non-negotiables
- Strong defaults
- The four tool types
- $fromAI(): how the agent fills tool parameters
- System prompt vs tool description
- Structured output: when and how
- Memory: brief mental model
- Binary and the agent boundary
- Human review (gate destructive tools)
- Chat agents (Slack, Discord, Teams, Telegram)
- RAG (retrieval augmented generation)
What does the n8n-agents skill do?
Design n8n AI agents, chains, classifiers, extractors, tool calling, memory, RAG, structured output, and human-review flows.
How do I install it?
Run `npx skills add sickn33/agentic-awesome-skills --skill n8n-agents --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From sickn33/agentic-awesome-skills, a repository with 44,414 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.