The insight most owners miss
Your result has two axes, not one
Most quizzes give you a single score. This one separates two things that are easy to confuse — how capable your setup is, and how much of it still depends on you being there.
RATE YOUR CLAUDE SETUP
How well you've set up & use Claude
Why you're doing this: a setup that serves the business, not just a faster you.
Each area shows what it means for your business as you go.
Maturity shape
By area
Illustrative of your configuration's risk of off-target output — not a measured model statistic.
A sample result. Strong in every area but one — a Dependency score of 30. That's a Power User who's still doing everything themselves. The skill builds this live as you answer.
Axis 1 · Capability archetype
How capable your setup is — names one of four:
- Explorer — Just chatting. Curious, getting value, plenty of headroom.
- Tinkerer — Building little habits — projects, saved prompts, the odd automation.
- Power User — Fast and fluent. Claude is woven through your day.
- Sprawler — Lots running — but spread out, undocumented, hard to hand over.
Axis 2 · Dependency badge
A separate read on how much it still needs you in the chair:
The work is good — but it mostly happens when you're the one driving.
The system keeps working when you step away — that's the goal.
Being faster in the chair is not the same as being removed from the work. Keeping these two apart is the whole point.
An education tool disguised as an assessment
Nine areas — and every question teaches
It scores each area and builds a live scorecard as you go. You don't need to know what any of these mean already — that's rather the point.
Skills & prompting
The difference between asking Claude every time and teaching it once so it answers the same way for everyone.
Context management
An overloaded setup is exactly how an agent writes the wrong offer into your ads.
Projects & memory
Where your hard-won instructions live — and whether they survive when the person who wrote them moves on.
Automations & agents
Whether work happens because someone prompted it, or because the system ran on its own.
Dashboards, Claude Code & local tools
The surfaces that let Claude do real work on your files and screens, not just talk about it.
MCP connectors
How Claude reaches your real tools — your inbox, your sheets, your CRM — safely or not at all.
CLI
The power-user layer. Worth knowing it exists, and whether it belongs in your setup yet.
Dependency
The big one: if you stepped away for two weeks, how much of this quietly stops?
Security
What Claude can see and touch — and whether a wrong click could send something it shouldn't.
Who it's for
Owners of high-revenue, low-headcount businesses who are already getting value from Claude — and want an honest read on where it's carrying the business and where it's quietly carrying you.
If you just chat
You'll see what you're getting right and the few moves that quietly compound.
If you tinker
You'll find the gaps between your habits and a setup that runs without you in the loop.
If you build
You'll get a clear read on dependency and security before they become a problem.
How to use it
Three steps. It runs inside the Claude desktop app — nothing else to install.
- 1
Download the skill
Grab the file below — it's a Claude skill bundle, and it runs right inside the Claude desktop app.
- 2
Add it to Claude as a skill
Add the downloaded skill in the Claude desktop app.
- 3
Start a chat and ask for your rating
Open a conversation and say "rate my Claude setup." It works best in Cowork — where it can read files you point it at, like an exported config — but a plain chat works fine too.
A live demo of how we treat your data
The skill itself is the proof. It works from what you tell it, it's honest about what it actually knows, and it sends nothing anywhere without your explicit yes.
Reads only what you share
It works from your answers and any files you point it at. It can’t see inside your Claude account — so when it doesn’t know, it asks rather than guesses.
Honest about confidence
It separates what it detected from what you told it, flags self-reported answers, and never invents a stat. Even the risk gauge is labelled illustrative, not a measured number.
Sends nothing without a yes
The shareable image carries zero business data. If you want the report sent to hmn.plus, it shows you the exact email first and waits for a clear yes.
That's how hmn.plus works in general: you get the system and stay in control of it. We never quietly do the work in the background.
See where you really stand — in about 15 minutes
Free, runs inside Claude desktop, and you walk away with a shareable scorecard. The kind of honest read that's easy to put off and hard to regret.
More human per hour. The system that helps owners step out of robotic work — never a done-for-you service.
Curious how this fits the bigger picture? Read about Claude for business