- Status
- Fixture
- Kata ID
- FIXUX-001
- Maintainer
- Kata template maintainer
- Last verified
- Not available. A fixture is never verified. No check of this recipe has been performed.
- Found a problem?
- Report it
The outcome
A new user of a fictional invoicing app reaches their first sent invoice using only the keyboard.
For: Freelancers who have never used the app, including people who navigate without a mouse.
Done when
- Every onboarding step can be reached and completed with the keyboard alone
- Three moderated sessions are written up, with what each participant tried
- One finding is fixed and the before and after are shown
You make: Interactive prototype, Usability findings. Outcome: Usable onboarding.
Showcase and provenance
Not available. No learner has made this outcome yet, so there is no authentic example to show.
Brief, inputs and constraints
Inputs
- Fictional product description. Source: Written in the prompt below; no real company or customer. Terms: Fictional, free to reuse.
- Journey mapping method. Source: nngroup/customer-journey-mapping. Terms: Read on the publisher's site under its terms.
Constraints
- Keyboard only
- Fictional data
Capabilities practised
- Design interaction
- Test usability
Stack, permissions and cost
| Part | Component | Version | Permissions | Cost |
|---|---|---|---|---|
| Harness | Any chat assistant in a browser | Not available. Not recorded; the fixture has never been run | Reads only what you paste into the chat | Not available. Not measured |
| Instructions | The two prompts on this page | Fixture text, 2026-11-20 | None | None |
| Tool | Penpot | Not available. Not recorded; the fixture has never been run | Read and write one practice project | Not available. Not checked for this fixture; see the publisher's current plans |
Stack recipes
- Browser prototype: HTML prototype in a browser
- Penpot design file: Penpot prototype file (penpot/prototyping)
Why this stack
A browser prototype and an open-source design tool keep the exercise free of installs and licences, and both can be tested with a keyboard.
Alternatives
- Figma through an MCP server. Brings design context into an agent, but needs an account and has not been run for the Dojo. Not tested for this recipe.
Agents and review gates
Designer
Drafts the journey map and the prototype screens from the brief.
Review gate: The learner checks every step against the done-when list before testing.
Note taker
Turns the learner's session notes into findings, one per observed problem.
Depends on: Designer
Review gate: The learner compares each finding with their own notes; nothing a participant did not do is added.
Setup and smoke test
- Open a new chat in the assistant you use.
- Open Penpot in a browser and create an empty practice project.
- Paste the first prompt and keep the answer next to the project.
Sign-in and access: Sign in to Penpot with your own account. The chat assistant needs no extra access.
Smoke test
- Do
- Tab through the first prototype screen from the address bar.
- Expect
- Focus reaches every control in reading order and is always visible.
- Result
- Not run Nobody has recorded running this check. It is not a pass.
Prompts
Copy a prompt into the assistant you use. Replace anything that does not fit your situation, and never paste private or personal data you are not allowed to share.
Draft the journey
Gets a first journey map and screen list you can check.
Turn session notes into findings
Keeps findings tied to what participants actually did.
Practice exercise
Moderated usability testing. From Using moderated usability testing (GOV.UK Service Manual).
Terms: Read on the publisher's site. GOV.UK guidance is published under the Open Government Licence v3.0 unless it says otherwise.
Execution and handoffs
- Designer: Draft the journey and screen list with the first prompt. Hand-off: The learner checks the draft and builds the prototype.
- Learner: Run three moderated sessions with keyboard-only navigation. Hand-off: Session notes go to the note taker.
- Note taker: Write findings from the notes with the second prompt. Hand-off: The learner fixes one finding and records the before and after.
Tests, failures and corrections
| Check | Expected | Result | Evidence |
|---|---|---|---|
| Keyboard-only pass through every onboarding step | Every control is reachable and operable; focus is always visible | Not run | Not run, so there is no evidence. This is not a pass. |
| Contrast check of text and focus indicators against WCAG 2.2 AA | No failure on any step | Not run | Not run, so there is no evidence. This is not a pass. |
Failures and corrections
No failures have been recorded, because no check has been run.
Adapt it
- The product becomes a patient appointment booking service. Health data rules apply; the prototype must use fictional patients only, and the findings need a privacy review.
- Participants use a screen reader. Headings, labels and live announcements need checking as well as focus order.
Unseen Randori
Design onboarding for a public benefit application you have not seen before.
Constraints
- Plain language at a reading age of about nine
- Keyboard and screen reader use
How it is judged: A reviewer tries the journey without help and records where they stopped.
Belt evidence
Belt stages are proposed practice stages, not issued credentials. Evidence below is what a reviewer would look for; it is not a record that anyone has shown it.
Evidence a reviewer looks for
- Findings from three moderated sessions, with what each participant tried (Peer review)
- Keyboard and contrast checks on every onboarding step (Automated check, against Web Content Accessibility Guidelines (WCAG))
Belt criteria
- Demonstrate: Fixes one finding and shows the before and after
Connections
- Foundations
- Used by
- No other kata names this one as a foundation yet.
- Professional review
- Not required for practice. Real work in a regulated field still follows that field’s rules.
Standards in scope
- Web Content Accessibility Guidelines (WCAG), 2.2. Applies to: Web content and prototypes a learner builds in practice. Review scope: Level A and AA success criteria on the practice prototype only; no conformance claim for any product
Sources
Penpot
Prototyping (opens the publisher's site in a new tab)
- Free
- Docs
- Beginner
Penpot explains how to connect boards into interactive user flows and preview them. Use any assistant to propose a flow, build it in your own file, then test the navigation and review the result. The guide is public; editor access and any optional integration depend on your setup and permissions.
Nielsen Norman Group
NN/g: Customer journey maps (opens the publisher's site in a new tab)
- Free
- Article
- Beginner
Explains when to create a customer journey map and how to base it on real data. It helps you tell the customer path from the user path. Read it before you map the buying side.
W3C
How to Meet WCAG (Quickref Reference) (opens the publisher's site in a new tab)
- Free
- Docs
- Intermediate
A filterable checklist of the WCAG 2.2 success criteria with techniques for each. Use it to write testable accessibility checks for your site.
GOV.UK Service Manual
Using moderated usability testing (opens the publisher's site in a new tab)
- Free
- Docs
- Intermediate
Explains how to run a usability test where a person watches someone use a prototype or service and asks questions. Directly useful for testing your website and mockups.
Downloads
Plain text files. Read them before you use them; nothing runs when you download one.
- agents.md: Instructions for the two agent roles, to put in an AGENTS.md or a project's custom instructions.
- done-when.md: The done-when checklist as a Markdown file you can tick off in your own notes.
Maintenance and caveats
Review cadence: Not reviewed; fixtures change only with the template
- This fixture exists to test the template. Do not treat its steps as checked guidance.
- Product names, versions and plans change; check every tool's current documentation before use.



