Secondary School Repair Router (eduKateSG) — O-Levels Reliability Under Load

TITLE: Secondary School Repair Router (eduKateSG) — O-Levels Reliability Under Load
SUBTITLE: Fence Learning Systems for Secondary English + E-Math + A-Math (Z0–Z2)
VERSION: CivOS Unified Spec v1.x
MODE: Almost-Code / WordPress paste-ready
SCOPE: Z0 learner + Z1 family + Z2 tuition/school; Sec 1–4; O-Levels
GOAL: P3 performance (transfer + speed + stability) under higher variance and stress
========================================================
0) LOCK BOX: WHAT SECONDARY / O-LEVELS REALLY ARE
========================================================
Secondary school is a load jump.
O-Levels is a transfer + stress reliability test:
- more abstraction
- longer papers
- higher coupling between topics
- higher penalty for prerequisite gaps
- increased emotional volatility
Therefore:
Success requires:
- stronger buffers
- faster fencing (Θ<1)
- enforceable routines (Λ<1)
- systematic inversion + timed training
========================================================
1) THE UNIFIED ROUTER (ONE SYSTEM, MULTIPLE SUBJECTS)
========================================================
System: Secondary-RepairRouter
Stack: EducationOS (capability regeneration) + FenceOS (boundary actuation)
Subjects: Secondary English, E-Math, A-Math (optionally Sciences)
Outputs: topic stability + timed performance + transfer reliability
Loop:
Detect → Classify → Fence → Repair → Verify → Transfer
========================================================
2) SECONDARY LOAD DIFFERENCES (WHY FENCE MATTERS MORE)
========================================================
Secondary has:
- higher topic coupling (one gap breaks many chapters)
- higher abstraction (algebraic reasoning, functions, proof-like steps)
- longer TTC consequences (one term lost can break the year)
- social/emotional coupling (identity, peers, burnout)
Rule:
Fence earlier than Primary.
Silent attrition is the main enemy.
========================================================
3) UNIVERSAL DASHBOARD (SAME METRICS, STRONGER THRESHOLDS)
========================================================
Metrics:
- TTC (to WA/EOY/O-Levels)
- Buffer (sleep, stress, routine, emotional stability)
- R (rate dominance) = are weak nodes shrinking weekly?
- Θ (fence timing) = did we respond in 24–72 hours?
- Λ (enforcement) = can we enforce boundaries same day?
- Coupling Index C (recommended for secondary)
Secondary Traffic States:
- Green: stable weekly improvement
- Amber: volatility or weak nodes not shrinking
- Red: timed collapse + avoidance + multiple topic failures
========================================================
4) Z0 ROUTER: FAST DIAGNOSTIC (SECONDARY EDITION)
========================================================
Step 1: Phase per subject (P0..P3):
- English P?
- E-Math P?
- A-Math P?
Step 2: Identify coupling source:
- prerequisite gap? routine? language drift? time collapse?
Step 3: Fence:
- truncate scope (one failing node)
- apply stop-loss timers
- rebuild buffers if low
- reduce coupling (contain failure to one lane)
Step 4: Stitch:
- smallest repair loop + retest fast
========================================================
5) SECONDARY ENGLISH ROUTER (O-LEVELS)
========================================================
English.P3 Definition:
- reads with precision under time
- answers comprehension exactly (inference + evidence)
- writes coherent essays with structure and control
- language accuracy supports clarity
Failure Modes (Top 6):
SE1 Evidence Weakness:
- answers vague, no quote/reference, wrong inference
SE2 Question Misread:
- misses command words; answers wrong focus
SE3 Summary / Synthesis Failure:
- cannot compress and select
SE4 Writing Structure Failure:
- weak argument/flow; unclear paragraphs
SE5 Language Drift:
- grammar errors obscure meaning
SE6 Time Collapse:
- incomplete sections, rushed writing
Sensors:
- misread count per paper
- evidence usage rate (points supported)
- paragraph clarity score
- error density per 100 words
- completion rate under time
Repair Loops:
SER1 Command-word bank + question parsing template
SER2 Evidence extraction drills (quote → inference → answer)
SER3 Summary compression drills (limit words, keep meaning)
SER4 Essay planning frames (thesis → points → examples)
SER5 Timed section micro-sets + review checklist
Fence:
- if TTC short: prioritize comprehension precision + timed completion
- stop-loss: if stuck > 90s on a question → mark, move, return
- buffer protection: sleep + reduce chaos before exams
========================================================
6) E-MATH ROUTER (O-LEVELS)
========================================================
E-Math.P3 Definition:
- chooses method quickly
- executes cleanly with low error rate
- handles unfamiliar contexts (transfer)
- finishes with time to check
Failure Modes (Top 6):
EM1 Algebra Drift:
- manipulation errors, factorization, expansion mistakes
EM2 Method Selection Failure:
- wrong chapter/tool chosen
EM3 Careless Error Dominance:
- arithmetic slips, copying errors
EM4 Word Problem Translation Failure:
- cannot model situations
EM5 Graph/Data Handling Weakness:
- graphs, interpretation, coordinates errors
EM6 Time Collapse:
- too slow, cannot complete paper
Sensors:
- first-step latency
- algebra error rate
- method selection accuracy
- time per mark trend
- “lost marks to carelessness” %
Repair Loops:
EMR1 Algebra micro-drills + spaced retest
EMR2 Inversion practice (random mixed sets)
EMR3 Timed micro-sets (marks-per-minute training)
EMR4 Error taxonomy checklist + forced checking routine
EMR5 Modeling template (Given/Find/Let/Equation/Check)
EMR6 Graph daily reps (small, consistent)
Fence:
- if R>1: stop new chapters; repair algebra base first
- stop-loss: 10 minutes stuck → see model solution + redo
- downgrade to restore correctness before increasing speed
========================================================
7) A-MATH ROUTER (O-LEVELS)
========================================================
A-Math.P3 Definition:
- can do algebra-heavy steps reliably
- understands functions/graphs
- executes differentiation/integration with accuracy
- solves multi-step problems under time
Failure Modes (Top 6):
AM1 Prerequisite Algebra Failure:
- factorization, indices, surds, logs weak
AM2 Function Concept Weakness:
- domain/range, transformations misunderstood
AM3 Differentiation Procedure Errors:
- rules known but applied wrongly under load
AM4 Integration Setup Failure:
- cannot choose method / limits / constants correctly
AM5 Multi-step Coupling Collapse:
- one early mistake destroys entire problem
AM6 Time Collapse + Panic:
- blanking, cannot start
Sensors:
- algebra step error density (per question)
- function sketch accuracy
- method selection for calculus questions
- multi-step “survival rate” (reach final answer?)
- time-to-first-step for hard questions
Repair Loops:
AMR1 Algebra spine repair (daily 15-min base set)
AMR2 Worked-example inversion (remove scaffolds, vary forms)
AMR3 Rule fluency micro-sets (speed + accuracy)
AMR4 Multi-step checkpointing (verify at each step)
AMR5 Mixed calculus transfer sets (randomized)
AMR6 Timed sections + post-mortem redo (stitching)
Fence:
- if TTC short: only repair algebra spine + top-weighted calculus patterns
- step-fence: mandatory checkpoints every 2–3 lines
- stop-loss: if stuck 8 minutes → mark, switch question, return later
- reduce coupling: isolate error types, prevent cascade
========================================================
8) Z1 FAMILY TOOLKIT (SECONDARY: BUFFER FIRST)
========================================================
Secondary reality:
- sleep and stress dominate performance variance
- routine is the main buffer
Parent/Guardian Fences:
- fixed study slot (short daily)
- device boundary (enforceable: Λ<1)
- conflict fence: process enforcement, not content teaching
- sleep fence before exams
- recovery fence after failures (no shame spiral)
Minimum Viable Routine:
- daily 30–45 min (alternating subjects)
- 2 timed sets/week
- 1 mixed transfer set/week
- 1 error log review/week (10 min)
========================================================
9) Z2 TUITION/SCHOOL ENGINE (TRANSFER + LOAD TRAINING)
========================================================
Z2 Weekly Cycle:
- Diagnose smallest failing node per subject
- Repair micro-sets (spaced)
- Timed micro-sets (marks-per-minute)
- Mixed sets (inversion)
- Simulation (every 2–4 weeks; more frequent as TTC shrinks)
- Retest within 7 days; within 72 hours if drift spikes
Non-negotiables:
- inversion tests (kills “chapter illusion”)
- timed training (kills exam collapse)
- redo protocol (stitching): redo weakest questions until stable
========================================================
10) FAILURE MODE TRACE (REQUIRED)
========================================================
Trace:
A-Math algebra spine weak → E-Math method confusion → English misread
→ coupling high + buffer low
→ timed collapse → avoidance loop
→ R>1 for 2 weeks
→ Θ>1 (late reaction)
→ multi-subject P2→P0 drift near exams
Fence Repair:
Detect R>1 + coupling high → truncate scope (algebra spine first)
→ enforce routine same day (Λ<1)
→ fence within 24h (Θ<1)
→ stitch: AMR1+AMR4, EMR2+EMR3, SER1+SER5
→ retest in 72h → R<1
→ rebuild buffers → restore P3 reliability
========================================================
11) WORDPRESS PASTE-READY PAGE STRUCTURE
========================================================
H1: Secondary Repair Router: How to Reach O-Level Reliability (English + E-Math + A-Math)
H2: Why Secondary Is a Load Jump (Coupling + Abstraction)
H2: The Universal Dashboard (TTC, Buffer, R, Θ, Λ, Coupling)
H2: English Router (Failure Modes → Sensors → Repairs)
H2: E-Math Router (Failure Modes → Sensors → Repairs)
H2: A-Math Router (Failure Modes → Sensors → Repairs)
H2: Parent Toolkit (Buffer-first fences)
H2: Tuition/School Engine (Transfer + Load Training)
H2: Failure Mode Trace (Short)
H2: Copy/Paste Weekly Plan
========================================================
12) COPY/PASTE WEEKLY PLAN (SEC 1–4)
========================================================
Daily (Mon–Fri):
- 30–45 min: 1 weak node
- 10 min: error log / redo 2 hardest questions
Twice/week:
- timed micro-set (marks-per-minute training)
Weekly:
- 1 mixed inversion set (all topics mixed)
- 10 min dashboard check (TTC/Buffer/R/Θ/Λ/C)
If TTC ≤ 30 days:
- stop new topics unless prerequisites stable
- simulations weekly
- strict sleep fence
- retest within 72h after any score drop
END SPEC
========================================================