Task scenario linter

A Claude skill that checks your usability test tasks for wording that gives the answer away, such as naming the button, before you run the test.

Study designOctober 12, 2026

Download the skill

Enter your details to download the skill file.

What it does

A usability task like "Click Reports to find last month's spend" gives the answer away by naming the button. You end up testing whether people can read, not whether they can find things.

You give it your usability tasks and a list of the product's menu and button names. It tells you which tasks use the product's own words, give step-by-step instructions, contain words like "easy" or "simply", or have no clear point where the participant knows they're done. Use it before you run a usability test.

What you get back

  • Every task that uses a product label, with the label it matched.
  • Instructions, priming words and hypothetical wording, quoted in place.
  • Tasks that depend on an earlier one, or that join two tasks with "and then".
  • A check for a clear end state, scored for the type of study you're running.
  • Severity on every finding, so you can fix the critical ones first.

Install it in Claude, ChatGPT, Copilot or Gemini

The same skill file works in all four. Pick your tool for the steps.

Install in Claude

  1. 1

    Download the skill

    Fill in the short form to get the .skill file.

  2. 2

    Add it to Claude

    In the Claude app go to Settings → Capabilities → Skills and upload the file. In Claude Code, run the npx command shown after the form, or drop the folder into your project's skills folder.

  3. 3

    Just ask

    Ask in plain language. Claude picks up the skill on its own when the request matches.

How to use it

Once it's installed, ask in plain language. Your AI follows the skill's method when the request matches.

Example prompt 1
Check these six tasks against our nav labels before the unmoderated test goes out.
Example prompt 2
We're testing the new Approvals page on purpose. Lint the tasks but allow that label.
Example prompt 3
Do any of these tasks tell people where to click?

What it won't do

It won't write your tasks for you, and it won't insist on a rewrite when naming a label is the point, as in a directed feature test or a study with specialists who use those terms. It checks wording, so whether a task is realistic for your participants is still your call.

Questions

What does it need from me?

Your tasks, as Markdown or text, and the product's navigation labels and button text, one per line or as JSON. The labels are what make the main check work, and a designer can usually list them quickly.

Does it work for tree tests and qualitative studies?

Yes. Set the study type and it adjusts: a missing end state is critical in a quantitative study, a warning in a qualitative one, and not checked in a tree test, where choosing a location is the end.

What if I need to use a product term on purpose?

Pass it as an allowed term and it won't be flagged. Everyday words that happen to be labels, like Home, Help or Search, are reported as warnings rather than critical findings.

More skills

See all skills