n8n and AI automation
n8n is a workflow automation engine often used to call LLM APIs, chain HTTP requests, and integrate SaaS, effectively an agent runtime with a GUI. Self-hosting shifts security responsibility to you: authentication, TLS, secrets, patching, and network exposure.
Official docs: n8n’s documentation is authoritative for versions and flags. This section summarizes common production patterns; always cross-check n8n Docs for your release.
When to self-host vs n8n Cloud
n8n states that self-hosting requires server, scaling, and security competence; mistakes can cause data loss, security issues, and downtime, and recommends n8n Cloud if you are not comfortable operating servers (Docker install intro).
Security mindset
- Workflows are code: Webhooks and schedules execute with credential access you granted.
- Secrets belong in n8n credentials: Encrypted at rest when
N8N_ENCRYPTION_KEYis managed correctly (deployment variables). - Expose only what you must: Put the editor behind SSO or VPN; isolate webhook URLs at the reverse proxy.
Articles
| Topic | Article |
|---|---|
| Install | Setup |
| Production URL | Deploy |
| Env & DB | Configuration |
| Hardening | Security |
| Patterns | Use cases |
| Lab | Simple Phishing Incident Triage |
Related CSN hub
- Agents overview: tool use, approvals, and sandboxing patterns that also apply to LLM workflows