n8n
A visual workflow automation tool whose value depends on reliable inputs, clear failure handling and a license that fits the intended use.
- Evidence
- Researched
- Model
- Source available
- License / terms
- Sustainable Use License; commercial options
Connect steps that already make sense.
An automation might take an approved inquiry, normalize its fields, create a task and notify a responsible person. The hard part is often deciding what counts as approved, who owns the next step and what happens when a system is unavailable.
n8n gives a visual place to express that workflow. This profile is based on documentation, not a claim that we have benchmarked or deployed it for the Lab.
Use the correct license label.
n8n describes its software as source available under the Sustainable Use License, with commercial restrictions. It does not describe that license as OSI open source. Internal workflows and selling access to an automation service are different use cases. Check the exact edition and intended use against the official guidance. n8n Sustainable Use License guidance.
Design the retry before the happy path.
| Failure | Question to answer |
|---|---|
| The destination times out | Can this step retry without creating a duplicate record? |
| A field is missing | Does the workflow stop, ask a person or discard the record? |
| Credentials expire | Who receives the failure notification? |
| The same event arrives twice | Which identifier prevents duplicate work? |
| An AI step gives an uncertain answer | What can proceed and what requires review? |
These questions remain even when the canvas makes the connection look simple.
Keep business rules outside a single canvas.
Retain a plain-language description of triggers, rules, data mappings and responsibility. Keep credentials in the appropriate secret store, not in exported examples. A workflow export is useful only if a successor can recreate its connections and understand its assumptions.
Budget the hosted service or self-hosting work, paid features where required, external API usage and maintenance. Every connected service can change independently.
When the extra system earns its place.
Consider it when several systems need a visible, maintained orchestration layer. Use a native integration or a small scheduled script when that is easier for the team to operate. Start with a non-critical workflow and deliberately trigger one failure before trusting it.
Our AI dependency guide explains how to keep the enduring rules available outside an assistant.
Sources and further reading
Primary documentation checked September 14, 2026. Product behavior and terms can change; follow the source for current details.
KEEP EXPLORING