DevOps Is Full of Hidden Prerequisites
I published a small interactive artifact today:
The idea came from thinking about adaptive education, but the thing I care about is much more practical: DevOps courses are full of hidden prerequisites.
A student can appear to finish a Kubernetes lab while understanding almost nothing about networking, selectors, ports, DNS, RBAC, or reconciliation. They copied the YAML. The resource exists. The lab looks green. But the actual learning is still missing.
That is the dangerous kind of progress.
The graph makes those hidden foundations explicit. Red nodes are skills the course often assumes but does not teach directly:
- shell fluency
- Git and PR workflow
- YAML
- HTTP and JSON
- ports and DNS
Those nodes feed everything else: Jenkins, Ansible, Docker, Kubernetes, AWS, Vault, observability, and GitHub Actions.
For AI-assisted DevOps education, this matters even more. AI can generate the YAML. It can produce a plausible pipeline. It can make broken infrastructure look professional. The student’s job is no longer to type every line manually. The student’s job is to understand enough to know whether the output makes sense.
That means diagnostics become more important than lectures.
Instead of asking “did the student complete the lab?”, the better question is:
Which prerequisite did this failure reveal?
If a Kubernetes Service cannot reach a Pod, the missing skill might be labels and selectors. Or ports. Or readiness. Or the mental model of Service-to-Pod routing. Rewatching the entire Kubernetes lecture is a blunt instrument. A mastery graph lets the teacher route the student to the exact missing piece.
This is not a platform yet. It is intentionally just a static graph.
That is enough for the first useful version: make the invisible gaps visible, especially the ones that quietly block students while pretending the lab is about something else.