01
Prepare a safe source packet
You narrow the task to one referral question, remove unnecessary identifiers, and label each source so claims can be checked later.
Made for Clinical Psychologists
Move From Raw Notes to Focused Differential Questions
Also useful for: people in Healthcare
45 min
Advanced
Healthcare
Build Professional Skills
Need to Compare Options
Organize raw assessment material, compare plausible explanations, expose missing or conflicting evidence, and build targeted follow-up questions without treating AI output as a diagnosis.
We'll use them to write your prompt.
Completed ✓
Make Your Play
Give us the situation. We’ll build the prompt.
Raw notes often mix observations, client reports, history, scores, and early impressions.
Your answers save as you type.
What you'll do
This is the plan inside your prompt. Add your details first. Then follow these steps with ChatGPT or Claude.
02
Separate facts from impressions
You ask the AI to extract only supplied information and sort it by evidence source.
03
Map the presentation over time
You organize onset, course, context, severity, and functional effects before considering explanations.
04
Generate provisional explanations
You create a deliberately broad set of plausible explanatory categories without selecting a diagnosis.
05
Compare supporting and conflicting details
You test each explanation against evidence that supports it, challenges it, or remains missing.
06
Check overlap and confounds
You examine whether timing, context, health factors, response patterns, or other influences could produce a similar presentation.
07
Run a safety and urgency check
You isolate supplied details that may require immediate direct assessment or action under your procedures.
08
Draft discriminating questions
You turn each important uncertainty into a question whose answer could strengthen, weaken, or redirect an explanation.
09
Prioritize the next assessment
You rank questions by urgency and information value, then place them in a usable interview order.
10
Complete your clinical review
You verify the final plan against source material, professional standards, and the limits of AI assistance.
Make Your Play
Move From Raw Notes to Focused Differential Questions
Tell us what the prompt needs to know
Start with the questions below. Each answer personalizes the playbook and builds the prompt you’ll use at the end.
Your situation
Describe the referral question, assessment stage, setting, and decision the next round of information needs to support.
0/10000
Paste or upload de-identified notes, history, observations, and collateral information, with sources and dates when available.
0/10000
Enter de-identified measure names, scores, response patterns, validity information, and any interpretation limits you already know.
0/10000
List missing records, uncertain timelines, conflicting sources, language or cultural considerations, and other limits that could affect interpretation.
0/10000
Select the stage that best describes your current assessment work.
This playbook was updated. Review your inputs before copying.
What you'll do
Check a step when you finish it.
01
Prepare a safe source packet
You narrow the task to one referral question, remove unnecessary identifiers, and label each source so claims can be checked later.
02
Separate facts from impressions
You ask the AI to extract only supplied information and sort it by evidence source.
03
Map the presentation over time
You organize onset, course, context, severity, and functional effects before considering explanations.
04
Generate provisional explanations
You create a deliberately broad set of plausible explanatory categories without selecting a diagnosis.
05
Compare supporting and conflicting details
You test each explanation against evidence that supports it, challenges it, or remains missing.
06
Check overlap and confounds
You examine whether timing, context, health factors, response patterns, or other influences could produce a similar presentation.
07
Run a safety and urgency check
You isolate supplied details that may require immediate direct assessment or action under your procedures.
08
Draft discriminating questions
You turn each important uncertainty into a question whose answer could strengthen, weaken, or redirect an explanation.
09
Prioritize the next assessment
You rank questions by urgency and information value, then place them in a usable interview order.
10
Complete your clinical review
You verify the final plan against source material, professional standards, and the limits of AI assistance.
Fill in the fields above to generate your prompt preview.
Assembled prompt
Beginner instructions
Start with one referral question and a limited set of de-identified notes. Review every extracted fact against the source before using the questions. If the output introduces a detail, score interpretation, or diagnostic claim you didn’t supply, remove it.
Advanced instructions
Run the sequence separately for competing symptom clusters when the case is dense. Add source dates and confidence limits, challenge the leading explanation with disconfirming evidence, and compare which questions have the highest information value. Apply your own standards for test interpretation, cultural formulation, risk assessment, documentation, and informed consent.
Use this when
Use this after an intake, record review, or early assessment session when several explanations remain plausible and you need a focused plan for gathering better evidence.
You'll need
- Clinical context
- Source material
- Supplied measure results
- Known limits
- Assessment stage
You'll get
- A source-checked presentation summary, 3 to 6 provisional explanations, an evidence comparison that includes conflicts and gaps, a relevant overlap check, and a prioritized set of non-duplicative questions
- Also reviewed urgent concerns, rejected unsupported inferences, and confirmed that the final document doesn't state a definitive diagnosis.
By the end of this playbook
You will have:
- A source-checked presentation summary, 3 to 6 provisional explanations, an evidence comparison that includes conflicts and gaps, a relevant overlap check, and a prioritized set of non-duplicative questions
- Also reviewed urgent concerns, rejected unsupported inferences, and confirmed that the final document doesn't state a definitive diagnosis.
Examples
Progress 1 of 10