SG Tutor/School Operations Panel v0.1 (Z1–Z2) — Cohort Triage + Remediation Routing + Drift Early Warning

TITLE: SG Tutor/School Operations Panel v0.1 (Z1–Z2) — Cohort Triage + Remediation Routing + Drift Early Warning
VERSION: CivOS Unified Spec v1.x
MODE: Almost-Code / LLM-first / WordPress paste-ready
OWNER: eduKateSG Fence Learning Systems
SCOPE: Z1 Family/Tutor + Z2 School/Center Operations (Primary → Secondary → JC)
GOAL: Prevent “silent attrition” by running EducationOS like a pipeline:
      Screen → Classify → Route → Repair → Re-test → Stabilise (P2/P3)
CANONICAL CLAIM:
- The biggest education failure is not visible collapse. It’s silent attrition:
  students drift from P2 to P1 over months, then crash to P0 at exam load.
- This panel makes drift measurable, fixes routing, and allocates limited time correctly.
========================================================
0) PANEL IDENTITY
========================================================
Panel: SG-OPS-PANEL-Z1Z2
Type: Operations + Triage + Routing + Early Warning
Users:
- Z1: Parents, Tutors, small centers
- Z2: Schools, Heads of Dept, year-level coordinators
Outputs:
- Cohort Phase Distribution (P0–P3)
- Drift Alerts (derivatives)
- Coupling Risk Map (upstream deficits causing downstream failures)
- Remediation Routing Plan (who gets what intervention)
- Time Allocation Plan (CH/ai-style scheduling)
========================================================
1) CORE CONCEPTS (LOCKED)
========================================================
- Phase P0–P3 is the operational truth (reliability under load).
- Coupling means “repair upstream first” (English→Math WP, Algebra→Physics, etc.).
- FenceOS is the stop-loss mechanism: Truncation + Stitching.
- Drift is diagnosed by derivatives, not just level:
  d(TR)/dt, d(GR)/dt, d(MR)/dt, d(Avoid)/dt.
========================================================
2) MINIMUM DATA TO RUN THE PANEL (LOW-FRICTION)
========================================================
Per student, per week (10 minutes max):
- Math: 1 timed mini-set (10–15 min) + error tags
- English: 1 RR test (10 min) + 1 paragraph (10 min)
- Science: DR (5 min) + 90s mechanism explain
- Humanities: 1 causality paragraph + evidence line
- TTC (weeks), HoursBuffer (rough), Volatility (low/med/high)
Optional:
- Full test scores when available
- One writing sample per month
========================================================
3) TRIAGE CLASSIFICATION (COHORT ROUTING)
========================================================
Cohort Buckets (operational):
Bucket A: P0 Emergency
- collapse visible; cannot complete tasks
Action:
- immediate truncation + rebuild buffer + 1–2 core nodes only
Bucket B: P1 Fragile
- drills ok, transfer fails; high VFR; speed unstable
Action:
- meaning/mechanism repair + variation ramps; prevent P2→P0 crash
Bucket C: P2 Stable
- reliable on standard mix
Action:
- maintain with mixed practice + quick patching (ERC rule)
Bucket D: P3 Advanced
- strong transfer + speed; can handle novelty
Action:
- enrichment allowed IF core stability preserved; use as peer-support nodes
========================================================
4) SILENT ATTRITION DETECTOR (DRIFT EARLY WARNING)
========================================================
Silent Attrition = P2 capability decaying quietly.
Drift Triggers (examples):
- TR decreases 2 weeks in a row (d(TR)/dt < 0)
- Grammar errors rise under timing (d(GR)/dt > 0)
- RR drops while word list grows (RR paradox)
- NIT triggers appear (English/Hum) after being absent
- Science MR drops while DR stable (memorise but can’t explain)
- AvoidSlope rising (increasing avoidance)
Drift Alert Levels:
- Green: stable
- Amber: early drift (intervene lightly)
- Red: pre-collapse (Fence actions now)
Fence Rule:
- Red alerts trigger truncation + targeted stitching within 7 days.
========================================================
5) COUPLING RISK MAP (WHY THE “WRONG SUBJECT” FAILS)
========================================================
High-weight coupling edges (default monitoring):
- ENG inference → Math word problems (H)
- Math graphs/algebra → Physics motion/electricity (VH)
- Science data graphs → Geography data interpretation (H)
- English paragraph coherence → Humanities essays (VH)
- Vocabulary retrieval → Writing output quality (VH)
Coupling Alert Trigger:
- downstream failures match upstream signatures
Action:
- reroute remediation upstream before adding more downstream practice
========================================================
6) REMEDIATION ROUTING (WHAT INTERVENTION EACH BUCKET GETS)
========================================================
Intervention Types (standardised):
I1: Core Node Repair (CN-R)
- fix one prerequisite node (e.g., fractions, inference, algebra hygiene)
I2: Transfer Training (TT)
- variation + mixed sets + explain-why checks
I3: Speed Stability Ladder (SSL)
- timed ramps; checkpoint habits; finish-ability training
I4: Coherence/Mechanism Repair (CMR)
- NIT / NIT-S coherence; 2-link chains; paragraph skeletons
I5: Buffer Rebuild (BR)
- reduce overload; routine; prevent Θ≥1 late fencing
Routing Table (default):
- P0 → {I5 + I1 + I4} (no full papers)
- P1 → {I1 + I2 + I4} (+ light I3)
- P2 → {I2 + quick I1 patches}
- P3 → {I2 + enrichment + peer-support}
========================================================
7) TIME ALLOCATION PLAN (CH/ai FOR COHORTS)
========================================================
Constraint:
- Teacher/tutor time is limited. Allocate where it prevents collapse.
CH/ai Cohort Allocation Rules:
- Prioritise Red drift alerts first (pre-collapse)
- Then P0
- Then P1 (highest ROI in preventing future collapse)
- P2 maintenance gets minimal but consistent checks
- P3 gets enrichment only if it does not steal core repair resources
Suggested weekly allocation (per 10 students):
- 40%: P1 fragile (prevent crash)
- 30%: P0 emergency (stabilise)
- 20%: P2 stable (maintenance checks)
- 10%: P3 advanced (extension + peer-support)
========================================================
8) SCHOOL/CENTER WEEKLY OPERATIONS LOOP (CLOSED LOOP)
========================================================
Weekly Loop:
1) Screen (mini-tests + samples)
2) Classify (Phase buckets + drift alerts)
3) Route (assign interventions I1–I5)
4) Repair (run sessions)
5) Re-test (same mini-tests)
6) Stabilise (lock P2; build toward P3)
7) Log (only minimal metrics; avoid admin overload)
========================================================
9) OPERATOR CHECKLISTS (READY-TO-USE)
========================================================
Checklist Z1 (Parent/Tutor):
- Did we run 1 writing paragraph this week?
- Did we run 1 timed math set?
- Did we detect ERC (repeat errors)?
- Any NIT triggers?
- TTC and backlog: are we in emergency mode?
Checklist Z2 (School/Center):
- Phase distribution stable?
- Red drift alerts addressed within 7 days?
- Coupling repairs routed upstream?
- Time allocation aligned to prevention (P1/P0)?
- Any cohort pockets showing brittleness (high variance)?
========================================================
10) REQUIRED SHORT FAILURE MODE TRACE (CANONICAL)
========================================================
Trace:
No screening → silent attrition → wrong subject remediation → backlog grows →
TTC shrinks → late fencing (Θ≥1) → multi-student P2→P0 exam shock
Fence Repair:
Weekly screening → drift alerts → upstream coupling repairs →
truncate breadth early → stitch core nodes → re-test → lock P2/P3
========================================================
END: SG Tutor/School Operations Panel v0.1
========================================================

WORDPRESS PASTE-READY PAGE STRUCTURE
H1: SG Tutor/School Operations Panel (Z1–Z2) — Prevent Silent Attrition
H2: Why schools/centers fail: drift, not visible collapse
H2: Minimum data to run the panel (10 minutes per student)
H2: Phase buckets (P0–P3) and what to do for each
H2: Drift early warning (derivatives, not vibes)
H2: Coupling risk map (repair upstream first)
H2: Remediation routing (I1–I5 intervention types)
H2: Time allocation plan (CH/ai cohort scheduling)
H2: Weekly closed-loop operations cycle
H2: Operator checklists (Z1 + Z2)
H2: Failure trace + repair trace