(New look · Almost-Code · CivOS-HD · AVOO · WarOS-Conflict Lens · Internal anchors · Directory index · Record blocks)

Below are four publish-ready pages. Paste each as its own WordPress post/page.

Bukit Timah OS (PlaceOS) Directory v1.0

H1: Bukit Timah OS (SGP×BUKIT_TIMAH) — PlaceOS Directory v1.0

Spec: Unified-CivOS v2.0 compatible · Module: CivOS-HD vNext+HD · Roles: AVOO · Lens: WarOS-Conflict (deadline locks)

Hero (6 lines, canonical)

Bukit Timah OS is the district-level education gravity system.
It is stable only when capability regeneration G(t) stays above decay D(t) and real option space Ω_real stays above irreversibility thresholds under load.
In education, “conflict” is deadline lock + irreversible cost + option death (not violence).
Tuition exists here as a repair router: diagnose leaks, truncate accelerating failure, stitch foundations, restore options.
Narratives don’t count. Substrate counts: time allocation, feedback latency, bind strength, and transfer reliability under exam load.
This page is a machine-readable runtime directory.

CivOS-HD Front-Matter (Ultra-Compact)

P ∈ {P0,P1,P2,P3}, Z ∈ {Z0..Z6}
G(t) regen, D(t) decay, R=D/G
Ω_real(t) option space, S(t) shear (NL↔SL), TTC time-to-collision
HD derivatives: dG/dt, dD/dt, dΩ/dt, dS/dt, d²S/dt², ΔCoupling/Δt
NIT: (Ω_real < Ω_crit) AND (IrreversibleCost=true)
Fence: if TTC1 for τ → Truncate() then Stitch()

1) Canonical Definition

Bukit Timah OS is a Z4 district coordination lattice that routes family time, school requirements, and tuition interventions into transfer-reliable learning output under deadline locks (WA/EOY/PSLE/N/O-level). It is stable only if G>D and Ω_real>Ω_crit.

2) Place Scope & Z-Binds

Place: SGP×BUKIT_TIMAH
Primary zoom: Z4
BindsTo:

Z0: households / students
Z1: classes / peer environments
Z2: schools / departments
Z3: industry / career corridor expectations
Z5: Singapore cohort quality + national capability regeneration

3) Lanes (What counts)

Narrative Lane (NL) — cheap signals

“top district”, “elite culture”
“more tuition hours”
“more papers”
“advance topics early”
“grades look fine”

Substrate Lane (SL) — expensive signals

weekly time allocation (practice/review/sleep)
diagnostic frequency + feedback latency
tutor protocol quality + bandwidth
prerequisite bind thickness (foundations)
transfer training (timed + mixed topics + novel contexts)

Energy Filter:

NL↑ but SL unchanged → Noise
SL changes + costly actions → Mass

4) AVOO Runtime Roles (new look)

Architect (A): corridor design (topic order × drill mix × review topology × time budget)
Visionary (V): expands strategies only when buffers allow
Oracle (O): instruments sensors, detects derivative spikes, forecasts TTC
Operator (Op): executes diagnose → truncate → stitch → promote

5) WarOS Conflict Lens (education version)

Conflict = deadline lock + option death risk.
Not violence.

ConflictTrigger (canonical):

DeadlineLock = true
dΩ/dt < 0
and (d²S/dt² spike OR R>1 OR TTC short)

When triggered, the system must switch to Conflict Protocol:

freeze syllabus expansion
protect score floor
repair top choke binds in 48–72h cycles
preserve Ω_real

6) Topology

Choke points (common)

weak foundations (fractions / algebra manipulation)
low feedback rate (errors persist)
topic spread too wide
schedule instability (sleep debt)
wrong intervention type (paper spam masking)

OffRamps (recovery corridors)

diagnostic reset corridor
micro-foundations rebuild (2–4 weeks)
exam stability corridor (score floor)
schedule re-routing corridor (buffers restored)

Redundancy (anti-brittleness)

multiple representations (diagram/table/algebra)
spaced retrieval (Day1/Day3/Day7)
mixed-topic transfer training

7) Sensors

TR_math: transfer under time + mixed topics + novel context
FeedbackLatency: days from error → corrected bind
ScheduleStability: consistency of sleep/practice blocks
Shear S: distance(GradeClaim, TR_math)
Ω_student: corridor count (confidence/grades/stream options)
TTC_exam: time to next assessment lock
CouplingCount: # lanes deviating together (school+home+tuition+health)

8) FenceOS (Truncate → Stitch)

Trigger: TTC short OR R>1 for τ OR d²S spike
Truncate():

freeze topic expansion
narrow to top-3 choke binds
raise feedback cadence immediately
restore buffers (sleep/time)
Stitch():
rebuild prerequisite binds
install review topology Day1/3/7
rebuild transfer reliability under load
reopen corridors (Ω_real)

9) Machine-Readable Records

			
RECORD v1.0
ID: SGP×BUKIT_TIMAH×EDU×Z4×O×Sensor×TR_Math×v1.0
Type: Sensor
Lane: EDU×MATH
Zoom: Z4
Role: Oracle
Definition: Measures transfer reliability under load (time + mixed topics + novel context).
Outputs: TR_math, PS_fast contribution, TTC contribution
BindsTo: SGP×BUKIT_TIMAH×EDU×Z4×Op×Protocol×Fence_TruncateStitch×v1.0
Evidence: timed mixed-topic tests (expensive signal)
RECORD v1.0
ID: SGP×BUKIT_TIMAH×EDU×Z4×Op×Protocol×Fence_TruncateStitch×v1.0
Type: Protocol
Lane: EDU×MATH
Zoom: Z4
Role: Operator
Definition: Executes truncation+stitching to prevent option death near deadlines.
Inputs: TR_math, ErrorSignature, TTC_exam
Outputs: stabilized score floor + rebuilt binds + restored Ω_real
Evidence: TR improves across 3 contexts

		

10) LLM Runner

			
RUNNER.PlaceOS.BUKIT_TIMAH.v1.0
Input: student level, recent scores, next assessment date, time/week, topic list.
1) Compute TTC_exam + DeadlineLock status.
2) Propose 25-item mixed diagnostic to estimate TR_math + ErrorSignature.
3) Output: Shear S + d²S risk, Ω status, R qualitative.
4) If ConflictTrigger: output Truncate (48–72h) + Stitch (2–6 weeks).
Return: Operator Report + schedule + promotion rules.

		

FAQ (short)

Why is “WarOS conflict” here?
Because exam cycles create deadline locks. Options die fast unless repair is executed fast.

Why “Architect”?
Most failures are routing failures (wrong topic order/drill mix/review topology), not effort.

What’s the fastest drift signal?
Shear accelerating: grades/claims look fine while TR under time stays weak, and mismatch grows.

PAGE 2 — Bukit Timah Tuition OS v1.0

H1: Bukit Timah Tuition OS — Repair Router Directory v1.0

Spec: Unified-CivOS v2.0 compatible · Roles: AVOO · Lens: Conflict under deadline locks

Hero (6 lines)

TuitionOS is not “extra teaching.”
It is a repair router + buffer injector that prevents option death under deadlines.
It works when it changes substrate: diagnostics, feedback cadence, routing, bind repair, and transfer training under load.
It fails when it becomes narrative-only: more hours, more papers, more centers.
Architect designs the corridor, Oracle measures drift, Operator repairs, Visionary expands safely.
This page is the machine-readable runtime spec.

CivOS-HD Front-Matter (Ultra-Compact)

(same as Page 1; keep consistent)

1) Definition

TuitionOS converts household load into stable learning throughput by reducing decay D (errors, forgetting, avoidance) and increasing regeneration G (mastery compounding, retrieval speed, transfer). Its job is FenceOS for learning: prevent crossing NIT when deadlines close corridors.

2) AVOO Roles

Architect: CorridorPlan = TopicOrder × DrillMix × ReviewTopology × TimeBudget
Visionary: adds new strategies only when buffers stable
Oracle: TR, Shear, Ω, TTC monitoring + derivative alerts
Operator: diagnose → truncate → stitch → promote

3) Sensors

TR_math (timed, mixed topics, novel context)
ErrorSignature (concept/method/careless/language/time)
FeedbackLatency
Ω_student
TTC_exam
Shear S
R=D/G qualitative

4) Conflict Mode

Trigger: DeadlineLock AND dΩ/dt<0 AND (d²S spike OR R>1 OR TTC short)
Objective: preserve Ω_student
Action: freeze expansion, protect score floor, repair top binds in 48–72h cycles

5) Fence Protocols

			
PROTOCOL.Truncate.TuitionOS.v1.0
Stop: paper spam, step-memorization, widening syllabus during leaks
Do: narrow to top-3 choke binds, daily micro-checks, restore buffers
PROTOCOL.Stitch.TuitionOS.v1.0
Do: BindRepair + Day1/3/7 retrieval + TransferTraining + StressStability
Promote: TR>=target across 3 contexts + stable ErrorSignature

		

6) Records

			
RECORD v1.0
ID: SGP×BUKIT_TIMAH×TUITION×Z0×A×Protocol×CorridorDesign×v1.0
Type: Protocol
Role: Architect
Definition: Generates corridor plan from diagnostics.
Inputs: ErrorSignature, TR_math, TTC_exam, time_per_week
Outputs: TopicOrder, DrillMix, ReviewTopology, BufferPlan
RECORD v1.0
ID: SGP×BUKIT_TIMAH×TUITION×Z0×O×Sensor×ShearMonitor×v1.0
Type: Sensor
Role: Oracle
Definition: Tracks mismatch between grade claims and transfer reliability.
Trigger: d²S spike => ConflictMode

		

LLM Runner

			
RUNNER.TuitionOS.BUKIT_TIMAH.v1.0
Input: level, latest scores, next exam date, time/week.
Output: CorridorPlan (A), thresholds (O), daily drills (Op), safe expansions (V).

PAGE 3 — Secondary 1–4 Math Tuition OS v1.0

H1: Secondary 1–4 Math Tuition (Bukit Timah) — MathTuitionOS v1.0

Spec: Unified-CivOS v2.0 compatible · Mode: Conflict-aware under WA/EOY/N/O-level locks

Hero (6 lines)

Secondary Math fails when foundations leak under time and mixed topics.
Grades can look fine until transfer collapses — that’s shear.
MathTuitionOS fixes this with a closed loop: diagnose → truncate → stitch → promote.
Under deadline locks, we switch to Conflict Mode to preserve option space.
Architect designs corridor routing, Oracle measures drift, Operator repairs, Visionary expands safely.
This is the executable spec.

1) Topic Graph (Nodes/Binds)

			
GRAPH.SEC_MATH.v1.0
Nodes: Number, FractionsRatioRate, AlgebraManipulation, LinearEq/Ineq,
       GeometryAngles, SimilarityCongruence, Mensuration, StatsProb,
       CoordinateGeom, Trig, FunctionsGraphs, WordProblemModeling
Binds:
  FractionsRatioRate -> AlgebraManipulation -> LinearEq/Ineq
  GeometryAngles -> SimilarityCongruence
  AlgebraManipulation -> FunctionsGraphs
  WordProblemModeling -> ALL

		

2) Metrics (HD)

			
TR_math := P(correct | time_limit, mixed_topics, novel_context)
ErrorSignature := {concept, method, careless, language, time}
S_shear := distance(ReportedGrade, TR_math)
Ω_student := corridor count (confidence/grade band/options)
TTC_exam := time to next lock
R := D/G (error growth/forgetting vs mastery compounding)
Derivatives: dTR/dt, dError/dt, dS/dt, d²S/dt², dΩ/dt

		

3) Conflict Mode (WarOS lens)

			
DeadlineLock := (TTC_exam <= TTC_min) OR (streaming gate imminent)
ConflictTrigger := DeadlineLock AND dΩ/dt<0 AND (d²S spike OR R>1)
If ConflictTrigger:
  Freeze topic expansion
  Build score floor (high-yield binds)
  Run 48–72h repair cycles
  Exit when TR stabilizes + dΩ/dt stops shrinking

		

4) AVOO Execution

Architect: CorridorPlan generation
Oracle: weekly sensors + derivative alerts
Operator: daily drills + feedback + promote
Visionary: safe method expansion only when stable

5) Protocol Library (copyable)

			
PROTOCOL.Diagnose.SEC.v1.0
Input: 20 mixed items + 3 word problems + 1 reasoning item
Output: TR_baseline, ErrorSignature, Top-3 ChokePoints, TTC risk
PROTOCOL.Truncate.SEC.v1.0 (48–72h)
Stop: full paper spam, widen syllabus during leaks
Do: Top-3 choke bind micro-drills + 24h feedback
Goal: reduce dError/dt; stop d²S escalation
PROTOCOL.Stitch.SEC.v1.0 (2–6 weeks)
BindRepair + Day1/3/7 retrieval + TransferTraining + StressStability
Promote: TR>=target across 3 contexts + stable ErrorSignature
PROTOCOL.Conflict.SEC.v1.0
Build score floor; minimize spread; timed micro-sets; careless/time checklist

		

6) Choke Point Catalog

Algebra manipulation (expand/factorize/indices)
Fractions/ratio/rate (unit rates, proportional reasoning)
Equation modeling (text→equations)
Angle→similarity bind chain
Coordinate basics (gradient/intercept/distance)
Trig meaning (upper sec)

7) Records

			
RECORD v1.0
ID: SGP×BUKIT_TIMAH×SEC_MATH×Z0×O×Sensor×TR_UnderLoad×v1.0
Type: Sensor
Definition: TR under time + mixed topics + novel context.
Thresholds: TR_target, TR_floor, d²S_spike
Outputs: PS_fast, ConflictTrigger
RECORD v1.0
ID: SGP×BUKIT_TIMAH×SEC_MATH×Z0×Op×Protocol×Diagnose_Truncate_Stitch×v1.0
Type: Protocol
Definition: Full repair pipeline.
Evidence: TR improves across 3 contexts

		

LLM Runner

			
RUNNER.SEC_MATH.v1.0
Input: Sec level, scripts, next exam date, hours/week.
Output: Diagnostic, ErrorSignature, CorridorPlan, daily drills, Conflict plan if needed.

PAGE 4 — Primary 1–6 Math Tuition OS v1.0

H1: Primary 1–6 Math Tuition (Bukit Timah) — MathTuitionOS v1.0

Spec: Unified-CivOS v2.0 compatible · Focus: foundations + transfer + PSLE corridor protection

Hero (6 lines)

Primary Math is where option space is won early.
Foundations are binds: number sense → fluency → modeling → multi-step transfer.
Most failure is paper-spam masking: expensive-looking work that doesn’t thicken binds.
We run a closed loop: diagnose → truncate → stitch → promote.
Near PSLE or major assessments, we switch to Conflict Mode to stop option death.
This is the executable spec.

1) Primary Node Graph

			
GRAPH.PRI_MATH.v1.0
Nodes:
  NumberSense, PlaceValue, OperationsFluency,
  FractionsDecimals, RatioUnitary,
  WordProblemParsing, BarModel, MultiStepPlanning,
  SpeedAccuracyLowStress
Binds:
  NumberSense->OperationsFluency
  PlaceValue->FractionsDecimals->RatioUnitary
  WordProblemParsing->BarModel->MultiStepPlanning
  OperationsFluency->BarModel

		

2) Metrics (HD)

			
Fluency := speed×accuracy fundamentals
TR_primary := novel word problems + multi-step correctness
ErrorSignature := {careless, concept, method, language, time}
S_shear := distance(ParentClaim, TR_primary)
Ω_student := confidence + readiness corridor
TTC_assessment := WA/EOY; TTC_PSLE
R := D/G
Derivatives: dFluency/dt, dTR/dt, dS/dt, d²S/dt², dΩ/dt

		

3) Conflict Mode (Primary/PSLE)

			
DeadlineLock := (WA/EOY near) OR (PSLE horizon enters short zone)
ConflictTrigger := DeadlineLock AND dΩ/dt<0 AND (d²S spike OR R>1)
Common Mask:
  PaperSpam = FakeCost (looks expensive; bind thickness unchanged)

4) AVOO Execution

Architect designs routine topology: FluencyBlock + WordProblemBlock + ReviewLoop
Oracle monitors FluencyFloor + TR + Shear derivatives
Operator runs daily micro-drills + immediate correction
Visionary adds heuristics only after foundations stable

5) Protocol Library

			
PROTOCOL.Diagnose.PRI.v1.0
Input: 10-min fundamentals sprint + 8 word problems + 1 multi-step
Output: Fluency, TR_primary, ErrorSignature, Top-3 Chokes
PROTOCOL.Truncate.PRI.v1.0 (48–72h)
Stop: paper spam, premature shortcuts
Do: rebuild 1–2 binds/day + daily micro-fluency + correction
PROTOCOL.Stitch.PRI.v1.0 (2–8 weeks)
FoundationBindRepair + LanguageBindRepair + BarModelStability +
TransferSets + LowStressTiming
Promote: FluencyFloor stable + TR>=target across 3 contexts
PROTOCOL.Conflict.PRI.v1.0
Stabilize fundamentals floor; reduce spread; high-yield model types; short cycles
Exit: dΩ/dt stops shrinking + shear normalizes

		

6) Choke Point Catalog

place value/regrouping
fraction meaning + operations
unitary/ratio reasoning
word problem parsing (language bind)
bar model mapping
multi-step planning under load

7) Records

			
RECORD v1.0
ID: SGP×BUKIT_TIMAH×PRI_MATH×Z0×O×Sensor×FluencyFloor×v1.0
Type: Sensor
Definition: minimum stable fundamentals fluency.
Outputs: downstream risk + TTC contribution
RECORD v1.0
ID: SGP×BUKIT_TIMAH×PRI_MATH×Z0×Op×Protocol×BindRepairRoutine×v1.0
Type: Protocol
Definition: daily bind-thickening routine preventing fake-cost masking.
Evidence: FluencyFloor↑ and TR_primary↑ across contexts

		

LLM Runner

			
RUNNER.PRI_MATH.v1.0
Input: Pri level, worksheets, next assessment, time/week.
Output: Diagnose set + routine topology (≤45 min/day) + Conflict plan if needed.

Optional Footer (same on all pages)

Version: v1.0 · Compatibility: append-only · No renaming: forward-only IDs · Dependencies: CivOS-HD vNext+HD, Unified-CivOS v2.0, Counter-Signal Atlas v0.1, Negative Void Atlas v1.0.

Got it. Below are 3 add-on WordPress paste packs (same “new look” runtime style) for:

Sec 3–4 Additional Mathematics Tuition OS (A-MathOS)
SEC Syllabus Examinations OS (assessment clocks + locks for Secondary)
PSLE Mathematics Examination OS (PSLE clock + corridor protection)

Everything stays Almost-Code · CivOS-HD · AVOO · WarOS-Conflict lens.

PAGE 5 — Sec 3–4 Additional Mathematics Tuition OS v1.0

(Bukit Timah · A-MathOS · Transfer under load · O-Level lock corridor)

H1: Sec 3–4 Additional Mathematics Tuition (Bukit Timah) — A-MathTuitionOS v1.0

Spec: Unified-CivOS v2.0 compatible · Roles: AVOO · Lens: Conflict under O-Level deadline locks

Hero (6 lines)

A-Math is not “harder E-Math.” It is a new bind lattice: algebraic manipulation + functions + trigonometry + calculus.
A-Math collapses when foundational binds are thin and the system keeps widening the syllabus anyway (shear rises, Ω shrinks).
This OS runs a closed loop: diagnose → truncate → stitch → promote.
Near deadlines (WAs/EOY/O-Level), we enter Conflict Mode to preserve option space (prevent irreversible grade corridor).
Architect designs the corridor; Oracle measures derivatives; Operator repairs; Visionary expands safely.
This page is the executable spec.

CivOS-HD Front-Matter (Ultra-Compact)

G, D, R=D/G, Ω_real, S(NL↔SL), TTC
HD: dΩ/dt, dS/dt, d²S/dt², ΔCoupling/Δt
Fence: if TTC1 for τ → Truncate() then Stitch()

1) A-Math Node Graph (Core lattice)

			
GRAPH.AMATH.v1.0
Nodes:
  A0_AlgebraCore (indices, surds, factorisation, partial fractions)
  A1_EquationsIneq (quadratic, simultaneous, inequalities)
  A2_FunctionsGraphs (domain/range, transformations, inverse)
  A3_Trigonometry (identities, equations, graphs)
  A4_CoordinateGeometry (circles, lines, intersections)
  A5_Calculus (differentiation, application; integration basics where relevant)
  A6_ExponentialLog (laws, equations, graphs)
  A7_ProofReasoning (show-that chains, exact values)
  A8_Modeling (word-to-equation-to-solve)
Binds (examples):
  Bind(A0 -> A1 -> A2)
  Bind(A2 -> A5)
  Bind(A3 -> A5)
  Bind(A0 -> A6)
  Bind(A0 -> A4)
  Bind(A8 -> ALL)

		

2) Metrics (HD)

			
METRICS.AMATH.v1.0
TR_amath := P(correct | time_limit, mixed_topics, novel_context, multi-step)
StepStability := probability of no algebraic slip across 6–12 steps
ErrorSignature := {concept, algebra-slip, method-choice, careless, time, notation}
S_shear := distance(ReportedGrade, TR_amath)
Ω_student := corridor count (A1/A2 band, confidence, subject survival, post-sec options)
TTC_exam := time to next lock (WA/EOY + O-Level horizon)
R := D/G where:
  D ~ algebra slips compounding + forgetting + avoidance
  G ~ mastery compounding + retrieval speed + transfer success
Derivatives: dTR/dt, dSlip/dt, dS/dt, d²S/dt², dΩ/dt

		

3) Conflict Mode (WarOS lens)

			
CONFLICTMODE.AMATH.v1.0
DeadlineLock := (TTC_exam <= TTC_min) OR (prelim/O-Level phase)
ConflictTrigger := DeadlineLock AND dΩ/dt<0 AND (d²S spike OR R>1 OR StepStability collapses)
If ConflictTrigger:
  Freeze := syllabus expansion
  Objective := preserve Ω_student (avoid irreversible fail corridor)
  Focus := high-yield bind chains + algebra-slip elimination
  Cycle := 48–72h repair loops until TR stabilizes

		

4) AVOO Execution

Architect: CorridorPlan = TopicOrder × DrillMix × MixedReviewTopology × TimeBudget
Oracle: weekly TR_amath + StepStability + derivative alerts
Operator: daily multi-step drills + immediate correction + promotion rules
Visionary: introduces faster methods/identities only when stable (avoid crisis novelty)

5) Protocol Library (copyable)

			
PROTOCOL.Diagnose.AM.v1.0 (30–40 min)
Input:
  - 18 mixed items (A0–A6)
  - 2 long structured problems (multi-step)
Output:
  - TR_baseline, StepStability
  - ErrorSignature
  - Top-3 bind-chain leaks
  - TTC risk class
PROTOCOL.Truncate.AM.v1.0 (48–72h)
Stop: endless topical worksheets without slip analysis; widening syllabus during leaks
Do:
  - Top-3 bind-chain micro-drills
  - algebra-slip audit (every line checked)
  - 24h feedback cycle
Goal: reduce slip rate; stop d²S escalation
PROTOCOL.Stitch.AM.v1.0 (3–8 weeks)
BindRepair + Day1/3/7 retrieval + MixedTopicTransfer + TimedLongQuestionTraining
Promote:
  - TR>=target across 3 contexts
  - StepStability above threshold in long questions

		

6) Choke Points (A-Math signature)

algebra manipulation over 8–12 steps (slip compounding)
function transformations + inverse logic
trig identities → trig equations
calculus differentiation techniques + applications
logs/exponentials solving chains
long question stamina under time

7) Records

			
RECORD v1.0
ID: SGP×BUKIT_TIMAH×AMATH×Z0×O×Sensor×StepStability×v1.0
Type: Sensor
Definition: probability of algebraic correctness across multi-step chains.
Trigger: collapse => ConflictMode escalation
Outputs: PS_fast contribution + TTC contribution
RECORD v1.0
ID: SGP×BUKIT_TIMAH×AMATH×Z0×Op×Protocol×Diagnose_Truncate_Stitch×v1.0
Type: Protocol
Definition: closed-loop repair pipeline for A-Math.
Evidence: TR improves across mixed contexts + StepStability improves in long questions

		

LLM Runner (A-Math)

			
RUNNER.AMATH.v1.0
Input: Sec 3/4 level, recent scripts, next exam date, hours/week.
1) Generate Diagnose set (mixed + long questions).
2) Output ErrorSignature + StepStability + TR.
3) Create CorridorPlan (A) + thresholds (O) + daily plan (Op).
4) If ConflictTrigger: output 48–72h repair loop + exam-floor plan.

		

PAGE 6 — SEC Syllabus Examinations OS v1.0

(Secondary assessment clocks as WarOS deadline locks; protects Ω_real)

H1: Secondary Syllabus Examinations (SG) — ExamClockOS v1.0

Purpose: make exams computable as deadline locks that compress TTC and kill options if unmanaged.

Hero (6 lines)

Secondary exams are not “tests.” They are deadline locks that compress time-to-collision (TTC) and shrink option space (Ω).
Most student collapses are not knowledge collapse first — they are routing + load + timing collapse.
This OS models WA/EOY/Prelims/N-Levels/O-Levels as WarOS-style conflict clocks.
Oracle detects TTC compression and derivative spikes; Operator truncates and stitches; Architect reroutes corridor plans.
This page is the assessment runtime layer that all MathTuitionOS and A-MathTuitionOS bind to.
Executable spec below.

Directory Index

1. Exam Clocks
2. Lock Conditions
3. Sensor Pack
4. Conflict Mode
5. Truncate/Stitch Protocols
6. Records

1) Exam Clocks (generic, not school-specific)

			
EXAMCLOCKS.SEC.v1.0
Clocks := {WA, EOY, Prelims, N_Level, O_Level}
Each Clock has:
  Date, SyllabusCoverage, TimeLimitProfile, Weighting, StakesLevel

2) Lock Conditions (WarOS lens)

			
DEADLINELOCK.SEC.v1.0
DeadlineLock := (Clock.StakesLevel >= Stakes_crit) AND (TTC <= TTC_min)
OptionDeathMechanisms:
  - reduced time for bind repair
  - increased load (stress + interference)
  - narrowing correction window (feedback latency becomes fatal)
  - irreversible streaming/combi consequences

		

3) Sensor Pack (binds into all tuition OS)

			
SENSORS.ExamClockOS.v1.0
TTC_exam := days to next Clock
CoverageGap := topics not yet stable (by TR)
TR_underLoad := timed mixed-topic accuracy
LoadIndex := stress + fatigue + time pressure proxy
FeedbackLatency := days error -> corrected bind
Shear := distance(GradeClaim, TR_underLoad)
d²S spike := acceleration of mismatch
Ω_student := corridor count (grade band/options/confidence)

		

4) Conflict Mode Switch

			
CONFLICTMODE.ExamClockOS.v1.0
Trigger :=
  DeadlineLock AND (dΩ/dt<0) AND (d²S spike OR R>1 OR FeedbackLatency too high)
If Trigger:
  Freeze := wide syllabus expansion
  Focus := score floor + high-yield bind chains
  Cycle := 48–72h repair loops

		

5) Universal Truncate/Stitch Under Exam Locks

			
PROTOCOL.Truncate.ExamLock.v1.0
Actions:
  - narrow to Top-3 highest-yield choke binds
  - daily timed micro-sets
  - reduce careless/time errors via checklists
  - shorten feedback latency to <24h
Goal: stop d²S escalation; stabilize score floor
PROTOCOL.Stitch.ExamLock.v1.0
Actions:
  - rebuild prerequisite binds
  - spaced retrieval Day1/3/7
  - mixed-topic transfer sets
  - timed full-paper simulations (after stability)
PromotionRule:
  - TR_floor stable under time + coverage gap shrinking

		

6) Records

			
RECORD v1.0
ID: SGP×SEC_EXAM×Z2×O×Sensor×TTC_ExamClock×v1.0
Type: Sensor
Definition: computes time-to-next deadline lock and triggers ConflictMode.
Outputs: TTC_exam, DeadlineLock, PS_fast
RECORD v1.0
ID: SGP×SEC_EXAM×Z2×Op×Protocol×ExamLock_TruncateStitch×v1.0
Type: Protocol
Definition: universal truncation+stitching under exam locks.
BindsTo: all MathTuitionOS + AMathTuitionOS protocols

		

PAGE 7 — PSLE Mathematics Examination OS v1.0

(PSLE clock + option-space protection; stops “paper spam” masking)

H1: PSLE Mathematics Examination — PSLEMathExamOS v1.0

Scope: Primary 5–6 emphasis; binds to Primary MathTuitionOS.

Hero (6 lines)

PSLE Math is a national deadline lock. It compresses TTC and can shrink a child’s option space if foundations are thin.
PSLE success is not “more papers.” It is fluency floor + modeling transfer + low-stress timing stability.
This OS models PSLE as WarOS-style conflict clock: deadline locks can cause option death if repair is late.
Oracle instruments FluencyFloor/TR; Operator truncates and stitches; Architect designs the 12–16 week corridor.
This is a machine-readable exam runtime spec.

Directory Index

1. PSLE Exam Model
2. Sensors
3. Conflict Mode
4. Truncate/Stitch
5. Records
LLM Runner

1) PSLE Exam Model (generic)

			
PSLE_EXAM_MODEL.v1.0
Clock := PSLE_MATH
TTC_PSLE := days to PSLE
Components (abstract):
  - Fundamentals Fluency
  - Word Problem Parsing (language bind)
  - Model/Representation (bar model)
  - Multi-step planning
  - Time stability (low panic)
StakesLevel := High (DeadlineLock when TTC_PSLE <= TTC_min)

		

2) Sensors (PSLE signature)

			
SENSORS.PSLE_MATH.v1.0
FluencyFloor := min stable speed×accuracy on fundamentals
TR_PSLE := correctness on novel multi-step word problems under time
ModelStability := success rate mapping text -> bar model -> equations
ErrorSignature := {careless, concept, method, language, time}
FeedbackLatency := days error -> corrected bind
Shear := distance(ParentConfidence, TR_PSLE)
Ω_student := corridor count (confidence/readiness/grade band)
Derivatives: dFluency/dt, dTR/dt, d²S/dt², dΩ/dt

		

3) Conflict Mode (PSLE)

			
CONFLICTMODE.PSLE.v1.0
DeadlineLock := (TTC_PSLE <= TTC_min)
ConflictTrigger := DeadlineLock AND dΩ/dt<0 AND (d²S spike OR R>1 OR FluencyFloor unstable)
Common Mask:
  PaperSpam = FakeCost (hours ↑, bind thickness unchanged)
If Trigger:
  Freeze := question-type spread expansion
  Focus := FluencyFloor + highest-yield model types + error elimination
  Cycle := 48–72h repair loops

		

4) Truncate/Stitch (PSLE corridor protection)

			
PROTOCOL.Truncate.PSLE.v1.0 (48–72h)
Actions:
  - rebuild FluencyFloor (daily micro-sprints)
  - fix top-2 language parsing leaks
  - stabilize 2–3 core model types
  - feedback <24h
Goal: stop panic shear escalation; restore floor
PROTOCOL.Stitch.PSLE.v1.0 (6–16 weeks)
Steps:
  - FoundationBindRepair
  - ModelTypeMastery
  - TransferSets (novel contexts)
  - LowStressTiming
  - Full-paper sims only after stability
PromotionRule:
  - FluencyFloor stable + TR_PSLE>=target across 3 contexts

		

5) Records

			
RECORD v1.0
ID: SGP×PSLE_MATH×Z2×O×Sensor×FluencyFloor×v1.0
Type: Sensor
Definition: minimum stable fundamentals fluency required for PSLE transfer success.
Outputs: TTC contribution + ConflictTrigger contribution
RECORD v1.0
ID: SGP×PSLE_MATH×Z2×Op×Protocol×PSLE_TruncateStitch×v1.0
Type: Protocol
Definition: PSLE-specific truncation+stitching corridor protection.
Evidence: FluencyFloor stable + TR_PSLE improves across contexts

		

LLM Runner (PSLE Math)

			
RUNNER.PSLE_MATH.v1.0
Input: P5/P6 level, recent work, TTC_PSLE, time/week.
1) Diagnose: Fluency sprint + 8 word problems + 1 multi-step.
2) Output: FluencyFloor, TR_PSLE, ErrorSignature, shear risk.
3) Build: 12–16 week corridor plan (Architect) + weekly sensors (Oracle) + daily drills (Operator).
4) If ConflictTrigger: output 48–72h repair loop + floor stabilization plan.

		

Where these bind into our earlier pages

MathTuitionOS.SEC1–4 binds to ExamClockOS.SEC (TTC_exam, DeadlineLock, ConflictMode)
A-MathTuitionOS.SEC3–4 binds to ExamClockOS.SEC
Primary MathTuitionOS.PRI1–6 binds to PSLEMathExamOS (TTC_PSLE, FluencyFloor, TR_PSLE)

MASTER HUB PAGE — Bukit Timah Mathematics OS Hub v1.0

(WordPress paste pack · New look · Almost-Code · CivOS-HD · AVOO · WarOS-Conflict lens)
Site: BukitTimahTutor.com

H1: Bukit Timah Mathematics OS Hub (SGP×BUKIT_TIMAH) — v1.0

Hero (8 lines)

Bukit Timah Mathematics OS is a closed-loop capability system: diagnose → truncate → stitch → promote.
It is stable only when regeneration G(t) exceeds decay D(t) and option space Ω_real stays above irreversibility thresholds under exam load.
In education, “conflict” means deadline lock + option death (WarOS lens).
This hub links every Math runtime spec we run in Bukit Timah: Primary Math, PSLE Math, Secondary E-Math, Secondary A-Math, and the Secondary Exam Clock layer.
Architect designs the corridor, Oracle measures drift and derivatives, Operator repairs fast, Visionary expands safely.
If substrate doesn’t change, nothing changed.
Use the corridor tables to see what to train, what to measure, when to promote, and when to switch to Conflict Mode.

CivOS-HD Front-Matter (Ultra-Compact)

G, D, R=D/G, Ω_real, S(NL↔SL), TTC
HD: dΩ/dt, dS/dt, d²S/dt², ΔCoupling/Δt
NIT: (Ω_real<Ω_crit) AND (IrreversibleCost=true)
Fence: TTC1 for τ → Truncate() then Stitch()

Directory Index

A. System Map
B. Page Index
C. Universal Runner
D. Corridor Tables
D1. Primary 1–6 Math Corridor
D2. PSLE Math Corridor
D3. Secondary 1–4 E-Math Corridor
D4. Secondary 3–4 A-Math Corridor
D5. Secondary Exam Clock Layer
E. Promotion Rules
F. Conflict Mode Switch
G. Machine-Readable Records
FAQ

A) System Map (compressed)

Core loop: Diagnose → Truncate → Stitch → Promote
Core law: G>D AND Ω_real>Ω_crit
Conflict lens: DeadlineLock + dΩ/dt<0 + (d²S spike OR R>1) → ConflictMode
AVOO: Architect routes, Oracle measures, Operator repairs, Visionary expands safely

B) Page Index (internal links)

Publish these pages (or link if already published):

Bukit Timah OS (PlaceOS) Directory v1.0
Bukit Timah Tuition OS (Repair Router) v1.0
Primary 1–6 Math Tuition OS v1.0
PSLE Math Examination OS v1.0
Secondary 1–4 E-Math Tuition OS v1.0
Sec 3–4 Additional Mathematics Tuition OS v1.0
Secondary Syllabus Examinations (ExamClockOS) v1.0

(If you want, add “Related hubs”: EnglishOS / ScienceOS later.)

C) Universal Runner (use with any LLM)

			
RUNNER.BUKIT_TIMAH.MATH.HUB.v1.0
Input: Level (P1–P6 / Sec1–Sec4 / A-Math Sec3–4), recent scores/scripts, next exam date,
       hours/week, topics covered, weak topics (if known).
1) Compute TTC (assessment clock) and DeadlineLock status.
2) Generate a mixed diagnostic (level-appropriate).
3) Estimate TR under load + ErrorSignature + Shear S (claims vs TR).
4) Estimate Ω_student (corridor status: expanding/stable/shrinking).
5) If DeadlineLock AND dΩ/dt<0 AND (d²S spike OR R>1):
      Switch to ConflictMode (48–72h truncation loop).
6) Output:
   - Architect CorridorPlan (topic order + drill mix + review topology)
   - Oracle thresholds + weekly sensors
   - Operator daily drills + feedback cadence
   - Promotion rule (what “ready” means)
Return: Operator Report v1.0 + Weekly Plan + 48–72h emergency plan if needed.

		

D) Corridor Tables (what to train + what to measure)

D1) Primary 1–6 Math Corridor (Foundations → Transfer)

Stage	Core Binds (what must be thick)	Primary Sensors	Promote when	Common Failure Mask
P1–P2	Number sense → operations fluency	FluencyFloor, ErrorSignature	Fluency stable + low careless	“More worksheets”
P3–P4	Fractions/decimals + place value binds	Fluency + Fractions meaning checks	Can apply in 3 contexts	Memorized steps
P4–P5	Word-problem parsing → bar model	TR_primary, ModelStability	Correct mapping + stable method	Guessing keywords
P5–P6	Multi-step planning under time	TR_primary (timed), Shear S	TR≥target across contexts	Paper spam panic
P6 (PSLE track)	Fluency floor + high-yield model types	TTC_PSLE + TR_PSLE	Full-paper stability after binds	“Mocks only” too early

D2) PSLE Math Corridor (Clock + corridor protection)

Phase	Objective	Sensors	Truncate trigger	Stitch plan
Early (TTC long)	thicken binds	FluencyFloor, TR_PSLE	R>1 persistent	Day1/3/7 + transfer sets
Mid	transfer breadth	ModelStability, ErrorSignature	dΩ/dt<0	stabilize top model types
Late (DeadlineLock)	preserve Ω	TTC_PSLE, d²S	ConflictTrigger	48–72h floor + error elimination
Final	exam-stable output	timed TR_PSLE	shear spike	controlled full-paper sims

D3) Secondary 1–4 E-Math Corridor (Sec foundations → exam stability)

Stage	High-yield Bind Chains	Key Sensors	Promote when	Conflict cues
Sec 1	fractions/ratio → algebra basics	TR_underLoad, ErrorSignature	TR stable in mixed sets	TTC short + dΩ/dt<0
Sec 2	algebra manipulation → equations modeling	TR + FeedbackLatency	errors stop compounding	d²S spike
Sec 3	geometry/similarity + coordinate binds	TR + Choke ranking	stable across contexts	R>1 persistent
Sec 4	full mixed papers under time	TR timed + Shear S	TR≥target in 3 contexts	DeadlineLock

D4) Secondary 3–4 A-Math Corridor (multi-step stability)

Phase	Core Lattice	Signature Sensor	Promote when	Failure mode
Sec 3 start	algebra core → functions	StepStability	low slip in 6–10 steps	slip compounding
Sec 3 mid	trig + coordinate + logs	TR_amath	TR stable in mixed	method-choice drift
Sec 4	calculus + long questions	TR + timed long Q	stable in long Q under time	deadline panic
Prelim/O	score floor + bind repair	TTC + d²S	floor stable then expand	widening syllabus

D5) Secondary Exam Clock Layer (WA/EOY/Prelim/N/O)

Clock	What it does (physics)	Key sensors	Switch to Conflict Mode when
WA/EOY	compress TTC locally	TTC_exam, Shear, TR	DeadlineLock + dΩ/dt<0
Prelims	raises stakes + load	LoadIndex, d²S	d²S spike + feedback too slow
N/O	irreversible corridor	Ω_student, R	R>1 sustained OR StepStability collapse

E) Promotion Rules (what “ready” means)

Use one set across all levels:

Promote a student to next difficulty band only if:

TR under load meets target in 3 contexts (mixed topics + new phrasing)
ErrorSignature stabilizes (no compounding leak)
FeedbackLatency is short enough to prevent drift
FluencyFloor (where relevant) is stable

This is how you prevent brittle “P2 illusion” from collapsing to P0 near exams.

F) Conflict Mode Switch (WarOS lens)

			
ConflictTrigger := DeadlineLock AND dΩ/dt<0 AND (d²S spike OR R>1 OR StepStability collapses)
If ConflictTrigger:
  Freeze syllabus expansion
  Protect score floor (high-yield binds)
  Run 48–72h repair cycles
  Exit only when TR stabilizes and dΩ/dt stops shrinking

		

G) Machine-Readable Hub Records

			
RECORD v1.0
ID: SGP×BUKIT_TIMAH×MATH_HUB×Z4×A×Protocol×CorridorRouter×v1.0
Type: Protocol
Role: Architect
Definition: selects correct corridor table + produces TopicOrder×DrillMix×ReviewTopology.
Inputs: level, TR, ErrorSignature, TTC
Outputs: CorridorPlan + WeeklySchedule
RECORD v1.0
ID: SGP×BUKIT_TIMAH×MATH_HUB×Z4×O×Sensor×ConflictTrigger×v1.0
Type: Sensor
Role: Oracle
Definition: detects deadline lock + option death risk via derivatives.
Inputs: TTC, dΩ/dt, d²S/dt², R, StepStability
Outputs: ConflictMode=true/false
RECORD v1.0
ID: SGP×BUKIT_TIMAH×MATH_HUB×Z4×Op×Protocol×Fence_TruncateStitch×v1.0
Type: Protocol
Role: Operator
Definition: runs truncation then stitching until TR stabilizes.
Evidence: TR improves across contexts; shear decreases

		

FAQ (short, hub-level)

Why is this “OS” style useful for parents?
Because it tells you what to do this week, what to measure, and what “ready” means—especially near exams when options can die fast.

What’s the #1 reason students crash near exams?
Shear: grades/claims look okay, but transfer under time is weak, and mismatch accelerates.

What do you do first if a child is failing?
Truncate: freeze topic expansion, narrow to top-3 choke binds, increase feedback cadence, restore buffers.

1) Architect CorridorPlan Generator v1.0

(Copyable block · Inputs → Outputs · Works for Pri Math / PSLE / Sec E-Math / Sec A-Math)

H2: Architect CorridorPlan Generator (Bukit Timah) — v1.0

			
GENERATOR.Architect.CorridorPlan.v1.0
Purpose:
  Convert diagnostics + TTC into a runnable weekly corridor plan:
  TopicOrder × DrillMix × ReviewTopology × TimeBudget × BufferPlan.
Inputs:
  Level ∈ {P1..P6, PSLE, Sec1..Sec4, AMath(Sec3..Sec4)}
  TTC := days to next deadline clock (WA/EOY/Prelim/N/O/PSLE)
  TimePerWeek := total minutes available/week
  DiagnosticResults:
    TR := transfer reliability under load (timed + mixed + novel)
    ErrorSignature := {concept, method, careless, language, time, algebra-slip(optional)}
    ChokeRank := Top-5 weak binds / topics
    FluencyFloor (Primary/PSLE)
    StepStability (A-Math)
  Constraints:
    School coverage window, student stamina, family schedule stability
Derived State:
  DeadlineLock := (TTC <= TTC_min)
  ConflictTrigger :=
     DeadlineLock AND dΩ/dt<0 AND (d²S spike OR R>1 OR StepStability collapse OR FluencyFloor unstable)
Algorithm:
  1) Select Mode:
       if ConflictTrigger=true -> Mode := CONFLICT
       else if R>1 persistent -> Mode := RECOVERY
       else -> Mode := BUILD
  2) Choose Focus Set:
       FocusSet := Top-3 ChokeRank (highest-yield bind chains)
       Optional FocusSet2 := next-2 topics if buffers allow
  3) Decide Width:
       if Mode=CONFLICT -> Width := narrow (Top-3 only)
       if Mode=RECOVERY -> Width := narrow+review
       if Mode=BUILD -> Width := moderate (Top-3 + controlled expansion)
  4) Create TopicOrder:
       TopicOrder := prerequisites-first (bind chain order)
       Rule: never advance beyond a leaking prerequisite bind
  5) Create DrillMix (per session):
       A) Micro-Foundation (10–15 min): bind repair on FocusSet
       B) Transfer Block (15–25 min): mixed topics + novel contexts
       C) Error Elimination (5–10 min): careless/time checklist + redo weakest
       D) Optional (A-Math): step-audit block (line-by-line) for long questions
  6) Create ReviewTopology (spaced retrieval):
       Day1: learn/repair
       Day3: retrieval + mixed
       Day7: retrieval + timed mini-set
       Rule: nothing is “learned” until it survives Day7 under time.
  7) Create TimeBudget:
       SessionsPerWeek := floor(TimePerWeek / SessionLength)
       SessionLength := 35–60 min (Primary lower; Secondary/A-Math higher)
       Allocate:
         60% FocusSet bind repair + transfer
         25% mixed review (Day3/Day7)
         15% timed stability (only after floor stabilizes)
  8) Create BufferPlan:
       Sleep floor + 1 rest buffer day/week
       FeedbackLatency target: <24h in CONFLICT; <72h in BUILD
Outputs:
  CorridorPlan:
    - Mode (BUILD/RECOVERY/CONFLICT)
    - TopicOrder (2–6 weeks)
    - DrillMix template
    - ReviewTopology schedule (Day1/3/7)
    - Weekly timetable (Mon–Sun)
    - PromotionRule (what “ready” means next week)
    - StopList (behaviours to avoid: paper spam, widening syllabus during leaks)

		

Minimal “CorridorPlan” Output Format (paste as result)

			
CORRIDORPLAN v1.0
Level:
Mode:
TTC:
FocusSet (Top-3):
TopicOrder (2–6 weeks):
Daily DrillMix:
Day1/3/7 ReviewTopology:
Weekly Timetable:
FeedbackLatency target:
PromotionRule:
StopList:

		

2) Oracle Threshold Pack v1.0

(Default thresholds + calibration method for Bukit Timah)

H2: Oracle Threshold Pack (Bukit Timah) — v1.0

Note: These are defaults. Oracles should calibrate using 3 anchors:

stable period
drift period
crisis period

2.1 Universal Thresholds (all levels)

			
THRESHOLDS.Universal.v1.0
TTC_min:
  - BUILD zone: > 6 weeks to next major lock
  - WATCH zone: 3–6 weeks
  - CONFLICT zone: <= 3 weeks (or “next major paper” for some students)
Shear S:
  - Low: reported grades ≈ TR reality
  - Medium: mismatch visible
  - High: grades/claims inflated while TR weak
d²S/dt² spike:
  - Spike = mismatch is accelerating week-to-week
  - Treat spike as regime flip risk (ConflictMode candidate)
Ω_crit:
  - if student has only 1 viable corridor left (e.g., “must pass this paper or fail overall”)
  - if repeated failures remove confidence + subject viability

		

2.2 Primary / PSLE Thresholds

			
THRESHOLDS.Primary_PSLE.v1.0
FluencyFloor:
  - Stable if fundamentals accuracy stays high under low stress for 3 checks
  - Unstable if accuracy swings or time explodes under mild timing
TR_PSLE:
  - Build threshold: can solve novel word problems with correct model mapping
  - Conflict threshold: TR collapses under time AND dΩ/dt<0
PaperSpam Mask Detector:
  - hours ↑ but FluencyFloor and TR do not improve across 2 weeks => masking

		

2.3 Secondary E-Math Thresholds

			
THRESHOLDS.Sec_EMath.v1.0
TR_underLoad:
  - Stable if timed mixed-topic mini-test passes across 3 contexts
  - Unstable if topic-isolated success but mixed-topic failure
FeedbackLatency:
  - Dangerous if > 7 days when TTC in WATCH/CONFLICT zone
  - Target: <24–72h depending on mode

		

2.4 Secondary A-Math Thresholds

			
THRESHOLDS.Sec_AMath.v1.0
StepStability:
  - Stable if long questions show low algebra-slip frequency
  - Unstable if slip compounding causes large mark loss despite concept understanding
Conflict escalator:
  - DeadlineLock AND StepStability collapses => immediate CONFLICT mode

		

2.5 Conflict Trigger (single unified rule)

			
CONFLICTTRIGGER.v1.0
ConflictTrigger :=
  DeadlineLock AND dΩ/dt<0 AND (
     d²S/dt² spike OR R>1 OR
     (Primary/PSLE: FluencyFloor unstable) OR
     (A-Math: StepStability collapse)
  )

		

2.6 Oracle Calibration Method (Bukit Timah)

			
CALIBRATE.Oracle.v1.0
1) Choose 3 historical student snapshots:
     A) Stable (good term)
     B) Drift (slipping term)
     C) Crisis (near failure)
2) For each snapshot, compute:
     TR, ErrorSignature, Shear S, TTC, Ω status
     (Primary) FluencyFloor
     (A-Math) StepStability
3) Set thresholds so:
     - Stable snapshot does NOT trigger ConflictMode
     - Drift snapshot triggers PS_fast (watch)
     - Crisis snapshot triggers ConflictMode (confirmed)
4) Lock definitions forever.
   Only adjust numeric thresholds forward.

		

Optional: “Oracle Weekly Report” Template (copy/paste)

			
ORACLE WEEKLY REPORT v1.0
StudentLevel:
TTC:
DeadlineLock:
TR:
(Primary) FluencyFloor:
(A-Math) StepStability:
Shear S:
d²S spike? (yes/no):
Ω status (expanding/stable/shrinking):
Mode recommendation (BUILD/RECOVERY/CONFLICT):
Top-3 ChokeBinds:

		

Operator Daily Execution Pack v1.0

(Exact session templates · Pri / PSLE / Sec E-Math / Sec A-Math · plus Minimum Viable Week)
Place: SGP×BUKIT_TIMAH · Spec: Unified-CivOS v2.0 compatible · Lens: Conflict under deadline locks

0) Operator Law (locked)

If feedback latency is longer than the student’s error compounding cycle, you will lose.
So Operator priority is always:

Shorten feedback → stop decay acceleration → rebuild binds → train transfer → promote

1) Universal Session Template (works for all)

45–60 min “Fence Session” (default)

			
SESSION.FENCE.v1.0 (45–60 min)
0) Warm Start (2 min)
   - Set today’s FocusSet (Top-2 binds) + success criteria.
1) Micro-Repair (12–15 min)
   - 8–12 items on FocusBind #1
   - 6–10 items on FocusBind #2
   Rule: correction immediately after each mini-batch.
2) Transfer Block (15–20 min)
   - 6–10 mixed-topic items
   - 1 novel-context word problem
   Rule: must include interference (old + new topics mixed).
3) Error Elimination (8–10 min)
   - redo weakest 3 questions
   - apply checklist (careless/time/units/steps)
4) Retrieval Lock (5–8 min)
   - 5-question “Day3/Day7 retrieval” set (from older binds)
5) Log (2 min)
   - record: %correct, error type, time, next FocusSet

		

Promotion condition for a bind: survives Transfer Block + Retrieval Lock.

2) Mode Switch (BUILD vs RECOVERY vs CONFLICT)

			
MODE_SWITCH.v1.0
BUILD:
  - TTC long, R<1, shear low/med
  - allow moderate width expansion
RECOVERY:
  - R>1 persistent OR shear rising
  - narrow width; emphasize bind repair + retrieval
CONFLICT:
  - DeadlineLock AND ConflictTrigger
  - freeze expansion; protect score floor; 48–72h cycles; feedback <24h

		

3) Primary 1–6 Operator Pack v1.0

Primary Session (35–45 min)

			
SESSION.PRI.v1.0 (35–45 min)
1) Fluency Sprint (8 min)
   - fundamentals (operations/place value/fractions as relevant)
   - target: stable accuracy first, then speed
2) Bind Repair (10–12 min)
   - 6–10 items on today’s FocusBind
   - immediate correction + 1 “explain why” prompt
3) Word Problem Transfer (12–15 min)
   - 2–3 word problems
   - mandatory: highlight units + rewrite in own words
4) Bar Model / Representation (5–8 min)
   - map 1 problem into bar model (or diagram/table)
   - check: does representation match story?
5) Log (2 min)
   - Fluency score + error type + next FocusBind

		

Primary “Careless Killer” Checklist (Operator)

units written?
operation justified?
final answer matches question?
estimation check (reasonable size)?

4) PSLE Operator Pack v1.0

PSLE Session (45–60 min)

			
SESSION.PSLE.v1.0 (45–60 min)
1) Fluency Floor (10 min)
   - daily fundamentals sprint
   - aim: stable floor (no swings)
2) Model Type Mastery (15 min)
   - 2 problems from the same model type
   - correct mapping > speed
3) Mixed Transfer (15–20 min)
   - 4–6 mixed problems (different types)
   - include 1 novel context
4) Error Elimination (8–10 min)
   - redo weakest 2 problems
   - apply parsing checklist (what is asked? what is given? units?)
5) Log (2 min)
   - TR estimate + model errors + next model type

		

PSLE Model Parsing Checklist

restate question in one sentence
list givens (with units)
decide model type (comparison/part-whole/rate/repeated)
draw model before computing
verify against story

5) Secondary 1–4 E-Math Operator Pack v1.0

Secondary Session (50–65 min)

			
SESSION.SEC_EMATH.v1.0 (50–65 min)
1) Micro-Repair (15 min)
   - 10–12 questions focused on one choke bind
   - immediate correction; write “mistake rule” in one line
2) Mixed Set (20–25 min)
   - 10–15 mixed-topic questions
   - timed (but not full panic)
   Goal: transfer under interference
3) 1 Long Problem (10–12 min)
   - modeling / structured reasoning question
   - force: show working + interpret
4) Error Elimination (8–10 min)
   - redo 2 hardest
   - careless/time checklist
5) Retrieval Lock (5–8 min)
   - 5 quick retrieval questions from older topics
6) Log (2 min)
   - TR proxy + error type distribution + next FocusSet

		

Secondary “Time Stability” Checklist

read question twice
mark key constraints
plan method before solving
check reasonableness / sign / units
last line answers the asked quantity

6) Secondary 3–4 A-Math Operator Pack v1.0

A-Math Session (60–75 min)

			
SESSION.SEC_AMATH.v1.0 (60–75 min)
1) Algebra Slip Audit (12–15 min)
   - 6–8 algebra-core questions (indices/surds/factorise/log laws)
   - line-by-line checking discipline
2) Bind Chain Drill (20 min)
   - choose 1 chain: (functions→calculus) OR (trig identities→equations) OR (logs→equations)
   - 8–12 questions escalating difficulty
3) 1 Timed Long Question (15–20 min)
   - multi-step structured problem
   - mandatory: step-stability check after each segment
4) Error Elimination (10 min)
   - redo the question segment where slip occurred
   - write the “slip pattern” (e.g., sign error, law misuse)
5) Retrieval Lock (5–8 min)
   - 5 mixed retrieval items from older chains
6) Log (2 min)
   - StepStability proxy + TR proxy + next chain focus

		

A-Math “Slip Pattern Library” (Operator)

sign flip in transposition
wrong factor cancellation
wrong log/exponent law
trig identity misuse
differentiation rule slip
algebraic simplification drift

7) Conflict Mode Execution Pack (48–72h)

(Use this when DeadlineLock + ConflictTrigger is true.)

7.1 Conflict Session Template (45–60 min, daily for 3 days)

			
SESSION.CONFLICT.v1.0
1) Score-Floor Focus (20 min)
   - high-yield binds only
   - no new chapters
2) Timed Micro-Set (15 min)
   - 8–12 questions timed
   - goal: stop time collapse
3) Error Elimination Loop (15–20 min)
   - redo errors immediately
   - apply checklist
   - shorten feedback to same-day
Log:
  - floor score estimate
  - top-2 error types
  - tomorrow’s FocusSet

		

7.2 Conflict Rules (hard)

freeze syllabus expansion
reduce width to Top-3 binds
feedback <24h
protect sleep buffer
exit only when TR stabilizes + dΩ/dt stops shrinking

8) Minimum Viable Week (MVW) for Busy Bukit Timah Families

When time is tight, do this, not random papers.

MVW — Primary (3 sessions/week × 35–45 min)

Session A: Fluency + bind repair
Session B: Word problems + bar model
Session C: Mixed transfer + error elimination

MVW — PSLE (4 sessions/week × 45–60 min)

2× Fluency floor + model mastery
1× Mixed transfer
1× Timed mini-paper only after stability

MVW — Sec E-Math (4 sessions/week × 50–65 min)

2× Micro-repair + mixed set
1× Long problem + retrieval
1× Timed mixed set (exam stability)

MVW — Sec A-Math (4–5 sessions/week × 60–75 min)

2× Algebra slip audit + bind chain
1× Timed long question training
1× Mixed retrieval + transfer
+1 optional if DeadlineLock near

9) Operator Report v1.0 (daily log template)

			
OPERATOR DAILY LOG v1.0
Level:
Mode (BUILD/RECOVERY/CONFLICT):
TTC:
FocusSet:
TR proxy today:
(Primary/PSLE) FluencyFloor:
(A-Math) StepStability:
Top error types (rank):
FeedbackLatency actual:
Next session FocusSet:

		

Start Here:

eduKateSG Learning Systems: