Diag n8n
When n8n grows past one screen, put the workflow estate in a hierarchy
n8n is excellent at the local job: build a workflow, connect nodes, test the path, and keep the runtime close to the automation.
The problem starts later. A team does not end up with one workflow. It ends up with lead intake, support triage, enrichment jobs, notifications, exports, retries, AI helpers, scheduled maintenance, and one-off experiments that quietly become production.
At that point, the question is no longer "what node comes next on this canvas?" The useful questions become:
- Which business capability owns this workflow?
- What other workflows or services must be healthy before it can run?
- Which credentials block it?
- Which workflows are entry points and which are sub-workflows?
- What checks prove the estate is ready?
That is not a single-canvas problem. It is a hierarchy and dependency problem.
Keep n8n as the editor
Diag n8n does not try to replace the n8n editor. The workflow graph still belongs in n8n because that is where the runtime semantics live.
Hood2 sits beside it as the operating map:
Operations
Automation Estate
Revenue Ops
Lead Intake
Receive Lead
Validate Lead
Enrich Company
Route Lead
Support
Ticket Triage
Receive Ticket
Classify Ticket
Lookup Customer
Draft Response
That structure gives the team a stable place for docs, ownership, contracts, fixture payloads, readiness checks, and dependencies.
Dependencies should be explicit
The value is not only the tree. The tree becomes useful when nodes can depend on each other.
A workflow tutorial might depend on local n8n being reachable. A NASA API example depends on that public API being reachable. An AI agent workflow depends on model credentials. A production support workflow might depend on customer lookup, classification, and escalation services.
In Hood2, those become named conditions and links instead of prose buried in a README.
First Workflow
depends on Quickstart:runnable
depends on Local Instance:local_n8n_ready
depends on NASA API:api_reachable
When something is blocked, the map can show what is blocking it.
The Diag n8n angle
Diag n8n is the technical wrapper around this model. It surfaces the n8n workflow estate as a browsable operating structure:
- workflow groups by business capability
- workflow fixtures and local runtime readiness
- Hood2 project nodes
- tutorial and dependency maps
- implementation gaps, such as commands that are modeled but not executable yet
The point is simple: n8n remains the workflow editor and runtime. Hood2 becomes the operational memory around the workflow estate.
That is how a pile of workflows becomes something a team can operate.