n8n nodes span three core automation domains that together power most production workloads. AI nodes — led by the AI Agent node serving as the central reasoning engine with LangChain‑powered decision logic — connect to OpenAI (GPT‑4o), Anthropic (Claude), Google Gemini, and vector stores for RAG pipelines. DevOps nodes — GitHub, Slack (30+ operations), PagerDuty, and Datadog — automate CI/CD alerts, incident management, and monitoring. Data nodes — Postgres (6 actions), MySQL (6 actions), HTTP Request (universal API connector), and 400+ integration nodes — power ETL pipelines, bidirectional CRM sync, and cross‑platform data flows. This guide catalogues each domain’s core nodes and canonical use‑case patterns. [1] [2]

What AI nodes does n8n provide for building autonomous agents, RAG pipelines, and LLM chains?

n8n’s AI ecosystem is built around the AI Agent node — the central reasoning engine that uses LangChain‑powered decision logic to determine which tools to call based on user input. It connects to sub‑nodes for a Language Model (OpenAI GPT‑4o, Anthropic Claude, Google Gemini, IONOS Cloud), Memory (Window Buffer, Postgres Chat, Redis), and Tools (HTTP Request, Code, sub‑workflow calls). [2] [6]

The @n8n/n8n-nodes-langchain package provides 70+ specialized AI nodes across ten categories: Agents, Chains (ChainLLM, ChainRetrievalQA), Chat Models (LmChatOpenAi, LmChatAnthropic), Embeddings, Vector Stores (Pinecone, Qdrant, PGVector, Supabase), Tools, Memory, Document Loaders, Text Splitters, and Output Parsers. The agent starts a reasoning loop: it analyzes the prompt, decides which tool to call, calls it, processes the response, and repeats until it produces a final answer — up to a configurable maxIterations (recommended: 5–8). A community implementation of the Anthropic Agent Skills pattern enables a single AI Agent to behave as a specialist across dozens of tasks using progressive disclosure with Postgres and native n8n nodes. For the complete RAG pipeline blueprint from document ingestion to citation, see the n8n AI Nodes reference.

How do DevOps nodes automate CI/CD alerts, incident management, and real‑time monitoring?

The DevOps node set — GitHub (20+ features: issues, releases, pull requests, workflow dispatch), Slack (30+ operations: channels, messages, reactions, files, users, plus Send and Wait for Response for human‑in‑the‑loop approvals), PagerDuty (5 resource types: incidents, incident notes, log entries, users, incident status), and Datadog (metrics, monitors, dashboards, logs) — forms a complete incident management stack. [7] [4]

The canonical DevOps pipeline chains six stages: a monitoring webhook (Datadog, Prometheus) → Code node normalises the alert payload into structured JSON → Switch node routes by severity (critical, warning, info) → HTTP Request or PagerDuty node creates an incident → Slack node notifies the on‑call channel → Send and Wait for Response pauses for acknowledgement, with a timeout branch that escalates to a manager. For real‑time infrastructure monitoring, Datadog’s community node provides native access to seven API resources directly within n8n workflows. The Slack node’s interactive Messages enable engineers to acknowledge, reassign, or resolve incidents by clicking buttons directly in Slack without leaving the conversation. For the complete six‑stage incident pipeline with escalation policies, see the n8n DevOps Nodes guide.

What data nodes power ETL pipelines, database sync, and cross‑platform data flows?

n8n’s data nodes provide full CRUD plus specialized operations for structured storage and universal API access. Postgres and MySQL each offer 6 actions — Select, Insert, Update, Delete, Upsert, Execute SQL — with parameterized queries using numbered placeholders. The HTTP Request node serves as the universal integration adapter: when n8n lacks a native node for a service, this node bridges the gap by calling any REST or GraphQL API with full method, header, and body control. [9] [10]

Beyond database and API nodes, the Set node reshapes JSON fields through manual mapping or JavaScript‑generated custom fields; the Merge node combines multiple data streams with four modes (Append, Combine, SQL Query, Choose Branch) and five join types; and the Code node executes custom JavaScript (Node.js 18) or Python for transformations that exceed built‑in node capabilities. The canonical ETL pipeline chains a Schedule Trigger → HTTP Request or database node (extract) → Set or Code node (transform) → database node with Upsert operation (load). For high‑volume pipelines, the Execute in Batches option on Postgres or MySQL nodes fetches records in chunks of 100–500 to control memory. For complete ETL patterns with bidirectional sync and AI‑assisted field mapping, see the n8n Database Nodes guide.

How does the HTTP Request node connect n8n to any API that lacks a native integration?

The HTTP Request node is the universal API adapter — it calls any REST or GraphQL endpoint using standard HTTP methods (GET, POST, PUT, PATCH, DELETE, HEAD, OPTIONS) with full control over Query Parameters, Headers, and Body configuration. It supports built‑in authentication (Basic Auth, Bearer Token, API Key, OAuth), cURL import for rapid prototyping, and a configurable timeout (default 300 seconds per node). The node parses JSON, XML, text, and binary responses automatically. [10] [11]

This node powers cross‑industry use cases: aggregating multi‑platform messages (GitHub commits, weather alerts, news feeds) into a unified Slack notification center; calling data enrichment APIs like Clearbit or Apollo to enrich CRM leads before routing; pulling data from services that n8n doesn’t have pre‑built integrations for — effectively expanding n8n’s reach beyond its 400+ native nodes to any service with an API. The community Better HTTP Request node extends this with configurable retry logic targeting specific HTTP status codes (default: 429, 500, 502, 503, 504), enabling up to 10 retry attempts with configurable delay. For the complete HTTP Request node configuration guide covering pagination, binary data, and proxy settings, see the Core n8n Nodes guide.

How do Slack, GitHub, and OpenAI nodes serve as the connective tissue across all use‑case domains?

Three nodes appear across every use‑case domain because they provide universal communication, trigger, and intelligence functions. Slack (30+ operations) delivers real‑time notifications regardless of domain — an AI agent sends its output to Slack, a DevOps pipeline posts incident alerts to the on‑call channel, a data pipeline reports ETL job status to the analytics channel. The GitHub node (20+ features) triggers workflows on repository events and dispatches actions programmatically. [4] [7]

OpenAI (via the OpenAI Chat Model sub‑node within the AI Agent) brings reasoning to any domain — scoring support tickets by urgency, classifying error logs in DevOps pipelines, normalising unstructured data in ETL flows. The pattern of connecting Slack + GitHub + OpenAI forms the backbone of many production workflows: a GitHub webhook triggers on a failed deployment, OpenAI classifies the failure severity, and Slack delivers the alert with a direct link to the failed execution. The n8n-nodes-elevenlabs community node extends this further by enabling AI voice generation directly within n8n workflows. For the complete AI agent orchestration guide, see the n8n AI Agents & LLM Orchestration guide.

How do community nodes extend n8n’s capabilities with 1,000+ additional integrations?

As of 2026, n8n supports over 400 native integrations, with the effective number of usable integrations exceeding 1,000+ services when community‑built nodes and HTTP Request‑based API connections are included. Community nodes — nearly 8,150 packages published on npm — represent a significant portion of new integrations, providing faster support for emerging tools. The n8n Pulse community dashboard tracks growth, templates, nodes, creators, and events across the ecosystem. [5] [12]

Community nodes are installed via Settings > Community Nodes > Install a community node, then entering the npm package name. Verified nodes appear with a shield icon. On self‑hosted n8n 2.x instances, a checksum‑based vetting system may block unverified packages — set N8N_UNVERIFIED_PACKAGES_ENABLED=true to bypass. Notable community nodes include ElevenLabs for AI voice generation, ShipStation for shipping label automation, and the Secure Webhook node for production‑grade webhook authentication with HMAC/JWT/IP policies. Security note: community nodes run with the same level of access as n8n itself — only install from trusted sources and review the code before deployment. For the complete community node installation and verification guide, see the n8n Nodes for Beginners Hub.