Get Better Answers Start the course

A prompt is tested, not admired.

Nell Garvey, who writes and runs the course

Write one prompt for a real job, run it against six awkward cases, and watch the score move.

Start the course

Nell Garvey in her workroom, arms folded, half smiling.

The answers are fine.

You use Claude every week for work. What comes back is fine, and not quite good enough to send without rewriting.

Most prompting advice is admired, not tested. You have read tips that sound clever. This course does not ask you to believe any of them. It marks your work, and shows you the marks.

“Nothing here rests on being told a prompt is good.”

Nell Garvey

One real job. Six awkward cases. A score that moves.

  1. You pick a real job, one of thirteen, from a customer enquiry or a job advert to a consultant’s proposal, a board update, a support reply or an incident report. You stay with it for the whole course.
  2. You write a prompt for that job, in the page, on top of the version you last saved.
  3. It runs against six cases, including a complaint, one the business genuinely cannot do, and one built to tempt the model into inventing a fact. Each answer is produced on its own, so one easy case cannot carry the rest.
  4. It is marked: mechanical checks first, then a grader that sees the brief, the case and the answer, but never your prompt. A rubric, not an opinion: four rules an answer must satisfy and eight it should.
  5. It is logged: version, case, pass or fail, and the score out of eight. Your log is the evidence the course worked, or did not.
Nell Garvey typing at her laptop in the workroom, marked pages stacked beside her.
Nell writes and runs the courses herself. She is the only person behind them.

Marked 12 September 2026, by the course’s own exercise

Same enquiry. Same rubric. Two prompts.

Version 0Version 1.1, written on top

Reply to this customer email.

  1. Gives it the briefYou are answering a customer enquiry for the business described in the brief below.
  2. Fences the factsUse only the facts in the brief.
  3. Names the gapIf the brief does not answer something the customer asked, say plainly that you will check and come back, and never invent a fact, a price or a timescale to fill the gap.
  4. Sets length and toneWrite 90 to 150 words of plain British English, warm but not gushing.
  5. Keeps their orderAnswer every question the customer actually asked, in the order they asked it.
  6. Ends on a next stepClose by telling the customer what happens next.

The caseThe one designed to tempt a model into inventing a price it was never given.

Version 0 what most people write

2of 8fail

Version 1.1 what the course builds

7of 8 pass Not run. Nothing is scored until it runs.

Graded by the same run of the course’s own exercise. The grader saw the brief, the case and the answer. Never the prompt.

Version 0 is one line. It scored 2 of 8, and failed.

Version 1.1 is written on top of it. While you write, the score does not move.

Then it runs, and a grader that never sees the prompt marks the answer.

It cannot be flattered. Nell Garvey

Nell’s hands marking a printed page in red pen.

What the marking actually checks

  • Every claim in the answer appears in the brief you supplied.
  • A gap in the brief is reported as a gap, not filled in. The case that tests this is the one most bare prompts fail.
  • Nothing is left as a placeholder. Checked mechanically, so it cannot be argued with.
  • An answer that reads as a draft, or explains that it is an AI writing a draft, is a fail.

“The output is the work, not a note about the work.”

Nell Garvey

She would rather tell you what is not finished than pretend it is.

Nell Garvey at her desk, chin on her hand, a marked script and her laptop in front of her.
Nell Garvey. She answers problems herself: the form on the privacy page reaches her directly.

Where this is up to

Open, and in private beta.

All the lessons are there and the marking runs end to end.

It has not yet had its adversarial audit, and the free run allowance is deliberately modest while the costs are still being watched.

How the course tries not to lie to you

  • Every quotation carries the page it came from and the date it was read, under the quote, not in a footnote.
  • A script re-fetches every quoted page on a schedule and fails when the sentence is gone. A stale course should be a job on someone’s list, not a surprise for a reader.
  • No screenshots of the interface. Lessons describe what a thing does and what to look for, because the buttons move monthly.
  • Two claims that other people write as fact are deliberately not made, because when it came to it, nobody had tested them.

Three courses, open side by side.

Start with the main one. The other two are open alongside it.

The main course

Better Answers from Claude

23 lessonsabout 150 minutes

For someone who already uses Claude for work every week. Eight modules and a short start-here, five to eight minutes a lesson. Every lesson ends in a run you can score.

  1. Start here
  2. Context
  3. Task
  4. Requirements
  5. Examples
  6. Evidence and testing
  7. Claude’s own features
  8. Capstone
  9. Next door: what Claude Code adds

You finish with one named, versioned prompt for a real job of your own, and a run log showing the new version beat the old.

Open alongside

Your Own Job, Scored

16 lessonsabout 105 minutes

Puts your own work into the marker.

You finish with your own job scored by the runner on two models, a rubric you have argued with three times, and a changelog with a model-moved line.

Stands on its own

Checking Claude’s Work

16 lessonsabout 101 minutes

Checking what a model wrote against its source before it goes out.

You finish with a checking prompt you can run on your own work, tested on drafts with known faults: what it caught, what it missed, and what it wrongly flagged.

The spine of the main course is one prompt, improved twenty times.

Lesson 0.2 has you write a bare prompt and run it. That score is the baseline every later lesson has to beat.

Your first run is the baseline. Everything after has to beat it.

By the capstone you repeat the whole method on a job of your own.

Better Answers from ClaudeCover sheet

Price
Free while the courses are in private beta. You make an account and start.
Account
It is how the course keeps your prompt versions, your runs and your scores. Without one there is nothing to mark and nothing to compare against.
Runs
One exercise is twelve model calls, six to answer and six to grade. Every account gets a capped number, enough to finish every exercise with room to retry and redo. At the ceiling, the runner refuses politely and says why.
Questions
The form on the privacy page reaches Nell directly.
Start the course

An account, then lesson 0.1: what Claude is, and is not.