Medical Scenario Gauntlet — Instructor Version

RTT/1‑Aligned • Zero Drift • Structural • Patient‑Safe#

Instructor Materials for the Tri‑Regime Medical Module#

Instructor Overview#

This gauntlet evaluates a student’s ability to run all six medical operators under realistic, multi‑snapshot, multi‑regime, drift‑heavy conditions:

1. MEDICAL_METADATA_OPERATOR
2. MEDICAL_DRIFT_OPERATOR
3. MEDICAL_SUBSTRATE_OPERATOR
4. MEDICAL_REGIME_OPERATOR
5. MEDICAL_CONTINUITY_OPERATOR
6. MEDICAL_SYNTHESIS_OPERATOR

Students work with synthetic medical scenarios modeled after:

• Cleveland Clinic (US clinical regime)
• Ping An Good Doctor (China AI‑augmented regime)
• NHS.uk (UK public‑health regime)

Each scenario includes:

• multi‑snapshot drift
• substrate variation
• translation instability
• AI‑generated overlays
• commercial interference
• regime shifts

Students must produce structural clarity, not medical advice.

Instructor Notes (Global)#

• Students must use structural reasoning only.
• Students must never copy or paraphrase real medical content.
• Students must treat Ping An as high‑drift, translation‑unstable, AI‑augmented, commercialized.
• Students must treat NHS as very low drift, clarity‑first, public‑health oriented.
• Students must treat Cleveland Clinic as clinical baseline.
• Students must always disclose uncertainty and escalation indicators.
• Students must never collapse regimes into a single narrative.
• Students must always identify drift sources before synthesizing.
• Students must never produce medical advice.

Scenario Structure#

Each scenario includes:

• Target condition (synthetic)
• Three regime snapshots (US, China, UK)
• Time‑series drift (early vs late snapshots)
• Substrate variation (html, mobile_html, pdf, image)
• Translation variance (for China snapshots)
• AI overlays (China snapshots)
• Commercial overlays (China snapshots)
• Public‑health framing (UK snapshots)

Students must run all six operators on each scenario.

SCENARIO 1 — Chest Discomfort Across Regimes#

Instructor Setup#

Provide students with:

• US snapshot (Cleveland Clinic‑like)

  • stable html
  • low noise
  • conservative framing
  • explicit escalation logic

• China snapshot (Ping An‑like)

  • mobile_html
  • translation drift
  • AI triage suggestions
  • commercial prompts
  • template drift

• UK snapshot (NHS‑like)

  • clarity‑first
  • standardized escalation rules
  • very low drift

Instructor Expectations#

Students must:

• identify substrate differences
• detect translation + AI drift in China snapshot
• detect template drift across China snapshots
• classify regimes correctly
• extract continuity kernel (stable symptoms, risks, red flags)
• produce a patient‑safe synthesis

Common Pitfalls#

• collapsing regimes
• including AI suggestions in continuity kernel
• missing translation drift
• missing commercial overlays
• failing to disclose uncertainty

SCENARIO 2 — Persistent Fatigue with Multilingual Drift#

Instructor Setup#

Provide:

• US snapshot: stable, clinical, low drift
• China snapshot:
  • two versions with different translations
  • AI‑generated symptom clustering
  • commercial telemedicine prompts
• UK snapshot: public‑health framing, low drift

Instructor Expectations#

Students must:

• identify translation drift between China snapshots
• identify AI‑generated instability
• classify regime shifts
• extract stable symptoms (fatigue, weakness, exertion intolerance)
• identify red flags (rapid deterioration, fainting, severe shortness of breath)
• produce drift‑bounded synthesis

Common Pitfalls#

• treating translation variants as separate medical facts
• failing to detect AI drift
• missing regime shift in China snapshots

SCENARIO 3 — Fever with Substrate Instability#

Instructor Setup#

Provide:

• US snapshot: html
• China snapshot: image‑embedded text + OCR variant
• UK snapshot: html, clarity‑first

Instructor Expectations#

Students must:

• detect OCR risk
• detect substrate instability
• identify missing accessibility layers
• extract stable red flags (persistent high fever, confusion, difficulty breathing)
• separate regime‑specific framing

Common Pitfalls#

• treating OCR artifacts as medical facts
• ignoring accessibility stability
• failing to identify substrate interference

SCENARIO 4 — Shortness of Breath with Commercial Drift#

Instructor Setup#

Provide:

• US snapshot: stable
• China snapshot:
  • commercial upsell blocks
  • AI triage overlays
  • mobile_html template drift
• UK snapshot: standardized escalation rules

Instructor Expectations#

Students must:

• detect commercial drift
• detect AI drift
• identify regime‑specific escalation logic
• extract continuity kernel (shortness of breath + red flags)
• produce patient‑safe synthesis

Common Pitfalls#

• including commercial prompts in synthesis
• missing AI‑generated instability
• collapsing escalation logic across regimes

SCENARIO 5 — Multi‑Regime, Multi‑Snapshot Capstone#

Instructor Setup#

Provide:

• 6–9 snapshots across all regimes
• mixed substrates
• translation variance
• AI overlays
• commercial layers
• template drift
• regime shifts

Instructor Expectations#

Students must:

• run all six operators
• produce full metadata, drift, substrate, regime, continuity, and synthesis packets
• identify stable vs unstable elements
• identify regime‑specific differences
• produce a drift‑bounded, patient‑safe synthesis

Common Pitfalls#

• failing to maintain operator separation
• collapsing regimes
• missing drift sources
• producing medical advice
• failing to disclose uncertainty

Instructor Rubric (50 Points)#

Mastery: 45–50
Proficient: 38–44
Developing: 30–37
Needs Support: 0–29

Instructor Closing Notes#

This gauntlet evaluates:

• drift literacy
• substrate literacy
• regime literacy
• continuity extraction
• patient‑safe synthesis
• cross‑regime reasoning
• uncertainty disclosure
• escalation awareness

It is the capstone assessment for the medical module.