Reply to this customer email.
2 of 8. Fail.
Get Better Answers
A prompt is tested, not admired.
Write a prompt for a real job, run it against six awkward cases, and watch the score move.
The same customer enquiry. The same rubric. Marked twice.
One. The proof
Two prompts. One enquiry. One rubric.
12 September 2026. The case is the one designed to tempt a model into inventing a price it was never given.
You are answering a customer enquiry for the business described in the brief below.
Use only the facts in the brief.
If 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.
Write 90 to 150 words of plain British English, warm but not gushing.
Answer every question the customer actually asked, in the order they asked it.
Close by telling the customer what happens next.
7 of 8. Pass.
Two. The method
One real job. Six cases. A score out of eight.
-
You pick a real job
One of 13, 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.
-
You write
A prompt for that job, in the page, on top of the version you last saved.
-
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.
-
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.
-
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.
Three. What is marked
It cannot be flattered.
The grader sees the brief, the case and the answer. It never sees your prompt.
- The brief
- The case
- The answer
The grader
Your prompt Never seen
A rubric, not an opinion: 4 rules an answer must satisfy and 8 it should.
- 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.
Nothing here rests on being told a prompt is good.
Nell Garvey
Four. The tutor
One person behind it.
Nell Garvey writes and runs the courses herself. She is the only person behind them.
She answers problems herself: the fastest way to reach her is the form on the privacy page.
The output is the work, not a note about the work.
Nell Garvey, on what counts as finished
Five. The course
Better Answers from Claude
For someone who already uses Claude for work every week and wants better results out of it. Every lesson ends in a run you can score.
- 23
- Lessons
- 5 to 8
- Minutes each
- About 150
- Minutes in all
The spine 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.
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. By the capstone you repeat the whole method on a job of your own.
- 0
Start here
- What Claude is, and is not
- Your first prompt, and the score that follows you
- 1
Context
- Give it what it needs
- Source material beats description
- Too much context
- 2
Task
- What, for whom, and why
- The decision the answer supports
- 3
Requirements
- Your definition of done
- Facts, assumptions, and the word unproven
- Rules you state and rules you enforce
- 4
Examples
- One example beats a paragraph of description
- The near miss
- 5
Evidence and testing
- Re-reads and receipts
- Telling a claim from a receipt
- What you have been building
- Versions, and writing down why
- Diagnosing a weak answer
- 6
Claude’s own features
- Projects, or context that stays
- Skills, or the prompt as a file
- Memory, search, and things that are documents
- Emphasis, and reading the page for your model
- 7
Capstone
- Your own job, end to end
- 8
Next door
- What Claude Code adds
Two more courses, open alongside it
Your Own Job, Scored (16 lessons, about 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.
Checking Claude’s Work (16 lessons, about 101 minutes) stands on its own and teaches 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.
Start with the main course. Make an account and start.
Six. What is not finished
Where this is up to.
Nell would rather tell you what is not finished than pretend it is.
- 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.
- 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.
Colophon
Nell Garvey.
Free while the courses are in private beta.
You make an account and start. Make an account and start.
An account 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.
- What running costs
- Marking is not free: one exercise is twelve model calls, six to answer and six to grade.
6 to answer. 6 to grade. 12 calls.
- The cap
- Every account gets a capped number of runs, enough to finish every exercise with room to retry and redo. The cap is a number on a settings page, not a promise, and the runner refuses politely and says why when an account or the course as a whole reaches its ceiling.
- If something goes wrong
- The fastest way to reach Nell is the form on the privacy page.
Get Better Answers, getbetteranswersfromyourprompts.com. Privacy page. Terms. Your Own Job, Scored. Checking Claude’s Work.