A practical prompting course
Better answers from Claude, and a way to prove it.
Most prompting advice is admired, not tested. This course marks your work. You write one prompt for a real job, run it against six awkward cases, and watch the score move as you improve it. 23 lessons, one idea each.
Version 0 · what most people write
Reply to this customer email.
Version 1.1 · what the course builds
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.
Both scores are real. The same customer enquiry, the same marking rubric, graded by the same run of the course’s own exercise on 12 September 2026. The case is the one designed to tempt a model into inventing a price it was never given.
A prompt is tested, not admired.
Every lesson ends in a run. Every run is scored. By the fifth module you have a log of your own versions, and the argument for each one is a number that moved.
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.
Six cases, not one
Each job has six inputs, including a complaint, one the business genuinely cannot do, and one built to tempt the model into inventing a fact.
A rubric, not an opinion
Four rules an answer must satisfy and eight it should. Some are checked by a machine, some by a grader that never sees your prompt.
You finish with a prompt
Named, versioned, with a run log showing the new version beat the old one. That is what finishing means here.
What the marking actually checks
- Every claim in the answer appears in the brief you supplied.
Checked by a grader that is shown the brief and the answer, and never your prompt.
- 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.
- The answer reads as a draft, or explains that it is an AI writing a draft.
This is a fail. The output is the work, not a note about the work.
Where this is up to
Open, and in private beta. All 23 lessons are here and the exercise marking runs end to end, so you can make an account and work through the course today. It has not yet had its adversarial audit, and the free run allowance is deliberately modest while the costs are still being watched. If something is wrong, the fastest way to tell me is the form on the privacy page, which reaches me directly.
Two more courses are open alongside this one. Your Own Job, Scored puts your own work into the marker. Checking Claude’s Work stands on its own and teaches checking what a model wrote against its source before it goes out. See all three courses.