GreyBrain School of AI
← All GreyBrain fieldnotes

GreyBrain Pulse · Published 17 Sept 2026

A complete OSCE station, built from nothing

The teacher · What’s possible · F2

Candidate sheet, actor script, weighted checklist and rating scale — the whole station, ready for the department to mark up.

Why this matters for your role

The student’s skill: ask for an explanation, work through it, then verify it against the taught source.

Your worked example

Teaching source: Random allocation reduces systematic baseline differences between groups. Blinding can reduce some measurement and performance biases. Neither guarantees that a result applies to every population.

Put it into practice

Create a short teaching analogy, a synthetic viva question and a worked answer grounded only in the source. Include a misconception check and a question the learner must answer independently.

Practise with a synthetic example. Do not upload identifiable patient information. Check every factual claim and source before considering any real-world use.

Clinician prompt

A complete OSCE station, built from nothing

Candidate sheet, actor script, weighted checklist and rating scale — the whole station, ready for the department to mark up.

Copy · paste into your AI

Build a complete OSCE station. This is a teaching exercise; my department will review it. Skill being assessed: [NAME IT] Level: [WHO IS BEING EXAMINED] Station length: [e.g. 7 minutes] Produce every document the station needs: 1. Candidate instructions — exactly what they read outside the door, no more. 2. Simulated patient script, including the opening line, what they volunteer, what they only reveal if asked, and how they respond if the candidate is abrupt. 3. Examiner checklist with weighted marks, written so two different examiners would score the same performance the same way. 4. A global rating scale with four levels described in observable behaviour. 5. The three things that will go wrong when this station is run for the first time, and how to prevent each. 6. What a borderline candidate looks like at this station, in three lines. End with a short section titled "Your check" listing what I must verify myself before using any of this.

Synthetic practice example. Run it as-is and check the output against the supplied facts.

Your check

Run it once with a colleague playing the candidate before you use it on students. Check the mark weighting reflects what your curriculum actually says matters.

Runs inClaude ChatGPT Copy the prompt first — these links open the tool, they do not carry the text.

A draft for departmental review. Assessment used for progression must be approved through your institution’s own process.

Sources inform this fieldnote; their publishers do not endorse GreyBrain. This exercise does not confer clinical competence or a certificate.