SG Student Diagnostic Prompt Pack v0.1 (LLM-run) — Phase + Failure Mode + Coupling Router + Weekly Plan

TITLE: SG Student Diagnostic Prompt Pack v0.1 (LLM-run) — Phase + Failure Mode + Coupling Router + Weekly Plan
VERSION: CivOS Unified Spec v1.x
MODE: Almost-Code / Copy-Paste Prompts / WordPress paste-ready
OWNER: eduKateSG Fence Learning Systems
SCOPE: Primary → Secondary → JC (Math/English/Science/Humanities)
GOAL: One LLM prompt that outputs: Phase scores, collapse risk, dominant failure mode, coupling causes,
      truncation + stitching plan, and a CH/ai weekly route plan.
SAFETY NOTE:
- This is an educational diagnostic tool. It does not replace medical or professional assessment.
- Use it to route learning repairs, not to label a child’s intelligence.
========================================================
0) HOW TO USE (1-MINUTE)
========================================================
Step 1: Copy “PROMPT-A (FAST DIAGNOSTIC)” into any LLM.
Step 2: Paste student evidence:
- last 2–3 tests (scores + topics wrong)
- 1 writing sample (English/Hum)
- 1 science structured answer sample
- time taken for each paper (if known)
- weeks to exam (TTC)
Step 3: LLM outputs:
- Phase per subject (P0–P3)
- CollapseRiskScore
- Dominant Failure Mode (FM-1..FM-6)
- Coupling causes (upstream deficits)
- TruncationPlan + StitchPlan
- Weekly CH/ai Route Plan
Step 4: Run PROMPT-B weekly (5 minutes) to update plan.
========================================================
1) REQUIRED INPUT TEMPLATE (PASTE THIS FIRST)
========================================================
INPUT TEMPLATE:
- Student Level: (P3/P4/P5/P6/Sec1/Sec2/Sec3/Sec4/JC1/JC2)
- TTC (weeks to next major exam): __
- Hours available per week (realistic): __
- Volatility this month (low/med/high): __ (CCA/events/illness)
- Math evidence:
  - Recent scores:
  - Topics failing:
  - Typical mistakes:
  - Time taken:
- English evidence:
  - Recent scores:
  - Writing sample (paste):
  - Typical mistakes:
  - Time taken:
- Science evidence:
  - Recent scores:
  - 1 structured answer sample (paste):
  - Topics failing:
- Humanities evidence:
  - Recent scores:
  - 1 paragraph/essay sample (paste):
  - Typical mistakes:
- Notes from parent/tutor:
  - Avoidance/complaints:
  - Confidence:
  - Homework behavior:
========================================================
2) OUTPUT SPEC (WHAT THE LLM MUST RETURN)
========================================================
OUTPUT FORMAT (STRICT):
A) Phase Scores:
- Math: P?
- English: P?
- Science: P?
- Humanities: P?
B) Collapse Risk:
- CollapseRiskScore: Low / Medium / High
- Why (top 3 drivers): {TTC/Buffer/Coupling/Backlog/NIT/Speed}
C) Dominant Failure Mode:
- FM-x: name
- Evidence markers
D) Coupling Causes (upstream-first):
- CouplingEdge1 (Weight VH/H/M): From → To
- CouplingEdge2 ...
- Repair priority order
E) Truncation Plan (what to stop doing for now):
- Stop/Delay list (max 5)
- Why (reduce Θ, rebuild buffer)
F) Stitch Plan (what to rebuild first):
- Top 5 nodes (with 1-line repair loop each)
G) Weekly CH/ai Route Plan (7 days):
- Session list with durations + tasks + re-test checkpoints
- Emergency mode if TTC short
H) Re-test Schedule:
- Weekly micro-tests
- Next full simulation date
========================================================
3) FAILURE MODE DEFINITIONS (LOCKED)
========================================================
FM-1 SpeedCollapse:
- time insufficient; accuracy drops under timing; SII high
FM-2 RetrievalCollapse:
- cannot pull vocab/formulas under time; RR/GRF low
FM-3 Meaning/CoherenceCollapse (NIT):
- contradictions, drift, incoherent paragraphs; NIT triggers
FM-4 MechanismCollapse (NIT-S):
- science explanations wrong; causal chain breaks; formula plugging
FM-5 CouplingMisrepair:
- fixing wrong lane; upstream deficit drives downstream failure
FM-6 BufferHollowing:
- too many topics; backlog high; TTC short; late fencing (Θ≥1)
========================================================
4) PROMPT-A (FAST DIAGNOSTIC) — COPY/PASTE
========================================================
PROMPT-A:
You are eduKateSG Fence Learning Systems running CivOS-HD diagnostics.
Your job: diagnose learning as a coupled lattice system across Math/English/Science/Humanities.
Use these locked concepts:
- Phase P0–P3 (reliability under timed load)
- Z0 learner level
- FenceOS: Truncation + Stitching
- Coupling: upstream deficits cause downstream failures
- NIT (English/Hum coherence) and NIT-S (Science mechanism coherence)
Given the student evidence below, produce the OUTPUT FORMAT (STRICT).
Rules:
1) Repair upstream before downstream when coupling evidence exists.
2) If TTC < 8 weeks AND backlog > 3 topics, trigger Emergency Mode (TruncationPlan required).
3) If NIT or NIT-S triggers, reduce complexity and rebuild coherence/mechanism before adding content.
4) If drills succeed but transfer fails, treat as false competence (P1 masquerading as P2).
Now process this input:
[PASTE INPUT TEMPLATE DATA HERE]
========================================================
5) PROMPT-B (WEEKLY UPDATE) — COPY/PASTE
========================================================
PROMPT-B:
You are updating last week’s plan.
Inputs: last week’s micro-test results + 1 new writing/science/hum paragraph.
Tasks:
- Update Phase per subject
- Update CollapseRiskScore and dominant failure mode
- Adjust TruncationPlan and StitchPlan
- Output a new 7-day route plan
Use the same OUTPUT FORMAT (STRICT).
Here is last week’s output summary:
[PASTE LAST WEEK OUTPUT SECTIONS A–H]
Here are this week’s new results:
- Micro-test results:
- Math timed set mistakes:
- English RR/grammar issues:
- NIT triggers (if any):
- Science mechanism sample:
- Humanities paragraph sample:
- Any schedule disruptions:
========================================================
6) PROMPT-C (SUBJECT-SPECIFIC DEEP DIVE) — COPY/PASTE
========================================================
PROMPT-C-MATH:
Diagnose Math only using SG-MATH Master Graph logic.
Return:
- Top 3 prerequisite nodes causing failure
- Speed vs transfer vs structure classification
- 5-day repair micro-plan + re-test
PROMPT-C-ENGLISH:
Diagnose English only using SG-ENGLISH Master Graph logic.
Return:
- RR/GR/CR/SR scores (qualitative ok)
- NIT triggers list + where it broke
- 5-day coherence-first repair plan
PROMPT-C-SCIENCE:
Diagnose Science only using SG-SCIENCE Master Graph logic.
Return:
- DR/MR/GRF/PR assessment
- NIT-S triggers + broken causal links
- 5-day definition→mechanism→application plan
PROMPT-C-HUM:
Diagnose Humanities only using SG-HUM Master Graph logic.
Return:
- Timeline/Causality/Evidence/Structure assessment
- NIT-ARG triggers + contradiction points
- 5-day thesis→evidence→judgement plan
========================================================
7) PROMPT-D (EXAM SIMULATION GENERATOR) — COPY/PASTE
========================================================
PROMPT-D:
Generate a 60–90 minute exam simulation block appropriate for the student’s level.
Requirements:
- Include 4 mini-sections: Math / English / Science / Humanities
- Each section must target 2 high-weight nodes and include one variation question
- Provide marking checklist + what failure signatures to look for
- Output also: how to classify dominant failure mode from this simulation
========================================================
8) OPTIONAL “PARENT-FRIENDLY EXPLANATION” PROMPT — COPY/PASTE
========================================================
PROMPT-E:
Explain the diagnosis to a parent in simple language without jargon.
Include:
- What is breaking (one sentence)
- Why it breaks (2–3 bullets)
- What we will do this week (3 bullets)
- What we will NOT do this week (truncation) and why
- What improvement should look like in 2 weeks
========================================================
9) WORDPRESS PASTE-READY PAGE STRUCTURE
========================================================
H1: SG Student Diagnostic Prompt Pack (LLM-run) — Fence Learning Systems
H2: How to use (1-minute)
H2: Input template
H2: Output format (strict)
H2: Failure Mode definitions (FM-1 to FM-6)
H2: Prompt A (Fast Diagnostic)
H2: Prompt B (Weekly Update)
H2: Prompt C (Subject Deep Dives)
H2: Prompt D (Exam Simulation Generator)
H2: Prompt E (Parent-friendly explanation)
========================================================
END: SG Student Diagnostic Prompt Pack v0.1
========================================================