Neurodivergent Curriculum Design

Workflow

1. Identify the learner goal, domain, and output format.
2. Lead with the real-world outcome before naming the concept.
3. Introduce an analogy before the formal definition.
4. Show the repeatable pattern in a table or other explicit structure.
5. Provide a fully worked example with each step visible.
6. Add one low-pressure practice task.
7. Add common errors or ambiguity handling when misinterpretation is likely.
8. End with a one-sentence pattern summary.

When the task is not a lesson but a dense record summary, adapt the same principles:

1. Start with the main reading view the user needs
2. Use a date-first timeline as the primary structure
3. Separate repeating categories into their own tables
4. Remove repeated carry-forward content unless it changed
5. Keep provider names, dates, tests, and meds traceable in each row
6. Add a short glossary if abbreviations or service-line labels repeat often

Lesson Structure

Use this sequence unless the user already supplied a house format:

1. Hook
2. Analogy
3. Concept in plain language
4. Pattern
5. Worked example
6. Practice
7. Common errors
8. Summary

Apply these rules:

• lead with outcome, not terminology
• define the term after the analogy, not before
• prefer a table over prose when teaching a repeatable pattern
• show all steps in examples; do not hide intermediate reasoning
• keep practice self-paced with no performance pressure language

S.T.A.C.K. Framework

Use S.T.A.C.K. when teaching prompt engineering or structured AI task design.

Teach S.T.A.C.K. as a checklist, not a rigid formula. Have learners fill it out before drafting the prompt itself. Review the filled checklist before producing the final prompt.

Analogy Standard

Default analogy pair:

• Driving: use for automation, prompting, agent workflows, API calls, or any procedural skill that becomes smoother after repetition
• Shoe-tying: use for skill acquisition that feels awkward at first but becomes automatic with practice

Example framings:

• You do not think about each step when you drive once the pattern is internalized. Prompting works the same way.
• The first time you tied your shoes, every step was deliberate. AI tools feel like that at first.

Use a domain-specific analogy only if it is clearer than the standard pair.

AuDHD Accommodations

Build these accommodations into the materials:

Use this ambiguity table format when needed:

Provider Or Professional Interaction Guides

Use a two-column response format for self-advocacy or professional interaction materials:

Keep the tone calm, factual, and self-advocating. Do not add apology language.

Medical Record Or Dense Timeline Summaries

When the user needs a long medical or life-record set organized for easier reading:

• make the primary view a date-first timeline
• keep one encounter or event per row
• use provider as a column, not the main grouping
• split repeating categories into separate views instead of repeating them in every encounter
• show only NEW, CHANGED, STOPPED, ABNORMAL, or REFERRED items when summarizing a visit
• keep a short glossary when abbreviating institutions such as UCH

Use this structure by default:

Use these compression rules:

• collapse repeated unchanged medications into a carry-forward statement only when necessary
• do not restate the same provider address or institutional boilerplate on every row
• prefer a one-line note summary over pasting full note text unless the user asks for full text
• if the same referral appears across multiple visits, keep one canonical referral row and reference later repeats as follow-up

Google Educator Or Community Context

When the context is GEG-OH, Grow with Google, or Google Educator programs:

• teach why before how
• use community-first framing
• prefer locally grounded examples when Cincinnati context helps
• shape outputs as module outlines, facilitation guides, accessibility audits, or professional learning materials as requested

Resources

When the user asks for handouts, lesson structures, or guides, prefer outputs that can be reused across modules instead of one-off prose explanations.

Validation

Before delivery, confirm:

1. every abstract concept receives a concrete analogy before formal definition
2. each section teaches one concept, not several at once
3. instructions use one action per bullet
4. pattern teaching appears before variation or edge cases
5. the output includes explicit headers, labels, and step numbers where relevant
6. any provider or professional interaction language is factual and self-advocating, not apologetic
7. any ambiguity-prone task includes a table that resolves interpretation
8. any record summary uses chronology as the primary reading order unless the user explicitly asked for another order
9. repeated facts are collapsed or moved to a secondary table instead of being restated across encounters
10. any abbreviations introduced for concision are supported by a short glossary

When the user requests a revision, preserve the teaching sequence and reduce friction, jargon, or ambiguity before adding new complexity.