eduKate Singapore

VIEW THIS AS

Auto mode follows the Route Engine until you choose a viewpoint.

YOU ARE HERE

ROUTE CHECK

CONNECTED TO

WHAT NEXT

Use the canonical route for this room, or HELP if you are unsure.

Next useful routeCanonical hubHELP

LanguageOS Universal Z0–Z3 Directory Table (Almost-Code)

Three learners review open books together at a classroom table, with stacks of textbooks, stationery and a whiteboard in the bright room.

1) LanguageOS Universal Z0–Z3 Directory Table (Almost-Code)

Copy/paste note: WordPress tables can break for LLM extraction. I’m giving this as a directory block list (machine-readable) instead of a visual table.

Start Here: 

Spec: LanguageOS.DirectoryTable.Z0_Z3.Universal
Version: v1.0.0
RecordType: DIRECTORY_ROW
Fields:
- row_id
- z_level
- skill
- failure_modes
- sensors (EL + VL)
- primary_repairs (EL then VL)
- example_routes
- drill (daily)

Z0 — Sentence Reliability (micro-control)

DIRECTORY_ROW: LANG.Z0.001 — Replace weak words

z_level: Z0
skill: weak_word_upgrade
failure_modes:
- vague adjectives ("nice", "good", "bad")
- filler intensifiers ("very", "really")
- vague nouns ("thing", "stuff")
sensors:
- VL.S1 SenseWrong
- VL.S5 IntensityMismatch
- EL.E6 LoadCollapse (filler stacking)
primary_repairs:
- EL: ER5 SplitLoad (remove filler)
- VL: R1 ReplaceByEdge + R4 TuneIntensity
example_routes: [ROUTE.MIC.309]
drill: "Upgrade 2 sentences/day; replace 2–4 tokens only."

DIRECTORY_ROW: LANG.Z0.002 — Fix collocations

z_level: Z0
skill: collocation_fix
failure_modes:
- unnatural pairings ("beautiful smell", "very suspense")
- incorrect verb-noun pairing ("do a mistake")
sensors:
- VL.S4 CollocationClash
- VL.S3 FrameBreak
primary_repairs:
- EL: ER5 SplitLoad (simplify clause)
- VL: R3 FixCollocation + R2 FixFrame
example_routes: [ROUTE.MIC.307]
drill: "Find 1 odd phrase/day; replace with 1 natural collocation."

DIRECTORY_ROW: LANG.Z0.003 — Insert link words correctly

z_level: Z0
skill: link_insertion
failure_modes:
- "then then then"
- random "therefore" with no reason
sensors:
- EL.E3 LinkFailure
- VL.CI.logic mismatch (connector wrong type)
primary_repairs:
- EL: ER3 InsertLinks (CAUSE/CONTRAST/ADD/CONCLUSION)
- VL: (support island) VL.CI.logic
example_routes: [ROUTE.MIC.306]
drill: "Take 2 sentences; add 1 link word that matches relation."

DIRECTORY_ROW: LANG.Z0.004 — Pronoun / reference clarity

z_level: Z0
skill: reference_clarity
failure_modes:
- "he/she/they" unclear
- missing actor noun
sensors:
- EL.E4 ReferenceLeak
primary_repairs:
- EL: ER1 NameActors (repeat key nouns)
- VL: VL.CI.clarity (rephrase for exact meaning)
example_routes: [ROUTE.MIC.305]
drill: "Highlight pronouns; replace 1–2 with nouns."

Z1 — One Paragraph Control (single idea, stable flow)

DIRECTORY_ROW: LANG.Z1.101 — PEEL paragraph (argument)

z_level: Z1
skill: peel_paragraph
failure_modes:
- claim with no support
- example with no explanation
- ends without link back
sensors:
- EL.E3 LinkFailure
- EL.E5 ScopeLeak
- VL.S1 SenseWrong (weak claim words)
primary_repairs:
- EL: ER3 InsertLinks + ER4 ShrinkScope
- VL: R1 ReplaceByEdge (claim precision) + VL.CI.evidence
example_routes: [ROUTE.ARG.101]
drill: "1 PEEL paragraph/day: 1 claim + 1 example + 1 mechanism + 1 link."

DIRECTORY_ROW: LANG.Z1.102 — Show-not-tell paragraph (narrative)

z_level: Z1
skill: show_not_tell
failure_modes:
- emotion labels only (“I was very scared”)
- sensory overload
sensors:
- VL.S5 IntensityMismatch
- VL.S4 CollocationClash
- EL.E6 LoadCollapse (adjective stacking)
primary_repairs:
- EL: ER5 SplitLoad
- VL: VL.CI.sensory + R4 TuneIntensity + R3 FixCollocation
example_routes: [ROUTE.NAR.017]
drill: "2 sensory details max; emotion in separate sentence."

DIRECTORY_ROW: LANG.Z1.103 — Explain-to-younger (definition clarity)

z_level: Z1
skill: explain_simple
failure_modes:
- vague definition
- no example
sensors:
- VL.S1 SenseWrong
- EL.E3 LinkFailure
primary_repairs:
- EL: ER3 InsertLinks (definition→example→why)
- VL: VL.CI.clarity + VL.CI.process
example_routes: [ROUTE.EXP.213]
drill: "Define 1 term/day + 1 small example."

Z2 — Two Paragraph Stability (story/argument with controlled scope)

DIRECTORY_ROW: LANG.Z2.201 — 2-paragraph suspense arc

z_level: Z2
skill: suspense_two_paragraph
failure_modes:
- jump to peak too early
- no resolution/clue
sensors:
- VL.S5 IntensityMismatch
- EL.E5 ScopeLeak
- EL.E3 LinkFailure
primary_repairs:
- EL: ER4 ShrinkScope (single arc) + ER3 InsertLinks
- VL: VL.CI.suspense + VL.CI.sensory_sound/sight + R4 TuneIntensity
example_routes: [ROUTE.NAR.002]
drill: "P1 build; P2 turn+resolve; max 2 collocations per paragraph."

DIRECTORY_ROW: LANG.Z2.202 — Regret + repair arc (relationship)

z_level: Z2
skill: regret_repair
failure_modes:
- blame mixed with apology
- lesson too broad (“life lesson”)
sensors:
- VL.S6 PolarityDrift
- EL.E5 ScopeLeak
primary_repairs:
- EL: ER4 ShrinkScope (1 rule)
- VL: VL.CI.ownership + VL.CI.apology + R6 ReducePolarity
example_routes: [ROUTE.NAR.003]
drill: "Write 1 mistake → 1 proof → 1 repair action → 1 rule."

DIRECTORY_ROW: LANG.Z2.203 — Calm control under time pressure

z_level: Z2
skill: perform_under_load
failure_modes:
- panic wording, messy steps
- no strategy shown
sensors:
- VL.S5 IntensityMismatch
- EL.E6 LoadCollapse
primary_repairs:
- EL: ER5 SplitLoad
- VL: VL.CI.calm_control + VL.CI.time_pressure + VL.CI.process
example_routes: [ROUTE.NAR.016]
drill: "1 internal monologue paragraph/day: worry → plan → action → outcome."

Z3 — Essay Logic (multi-reason, evidence-safe, tone-stable)

DIRECTORY_ROW: LANG.Z3.301 — Concession + rebuttal paragraph

z_level: Z3
skill: concession_rebuttal
failure_modes:
- strawman rebuttal
- binary good/bad language
sensors:
- EL.E5 ScopeLeak
- VL.S1 SenseWrong
- Z0.WEAK absolutes
primary_repairs:
- EL: ER4 ShrinkScope + ER3 InsertLinks (however/therefore)
- VL: VL.CI.politeness + VL.CI.argument + R1 ReplaceByEdge
example_routes: [ROUTE.ARG.102]
drill: "Concede 1 benefit; rebut 1 risk; conclude 1 rule."

DIRECTORY_ROW: LANG.Z3.302 — Evidence-safe hedging (broad topics)

z_level: Z3
skill: hedged_claims
failure_modes:
- “proves”, “always”, “everyone”
- missing limitation sentence
sensors:
- VL.S1 SenseWrong (prove vs suggest)
- EL.E5 ScopeLeak
primary_repairs:
- EL: ER4 ShrinkScope + add limitation node
- VL: VL.CI.uncertainty_hedging + VL.CI.neutrality
example_routes: [ROUTE.ARG.118]
drill: "Write claim + evidence + ‘likely because’ + limitation line."

DIRECTORY_ROW: LANG.Z3.303 — Criteria-based fairness judgement

z_level: Z3
skill: criteria_judgement
failure_modes:
- “unfair” with no criteria
- emotional overload
sensors:
- VL.S1 SenseWrong
- EL.E3 LinkFailure
primary_repairs:
- EL: ER3 InsertLinks (criteria→evaluation→conclusion)
- VL: VL.CI.fairness + VL.CI.neutrality
example_routes: [ROUTE.ARG.115]
drill: "Pick 2 criteria; judge 1 policy; conclude 1 rule."

2) VL↔EL Crosswalk (Which VL islands best repair each EL sensor)

Spec: LanguageOS.Crosswalk.EL_to_VL
Version: v1.0.0
RecordType: EL_SENSOR_MAP
Fields:
- el_sensor
- what_it_means
- best_vl_islands
- typical_vl_tripwires
- best_repairs (EL then VL)

EL_SENSOR_MAP: EL.E1 RoleConfusion

what_it_means: actor/action/object unclear (“who did what to whom”)
best_vl_islands: [VL.CI.clarity, VL.CI.process]
typical_vl_tripwires: [S1 SenseWrong (vague verbs)]
best_repairs:
- EL: ER1 NameActors + simplify sentence
- VL: R1 ReplaceByEdge (use exact verbs: admit/insist/notice/avoid)

EL_SENSOR_MAP: EL.E2 TimeDrift

what_it_means: tense/time inconsistent in narrative/argument
best_vl_islands: [VL.CI.time_pressure, VL.CI.process]
typical_vl_tripwires: [S3 FrameBreak (tense errors)]
best_repairs:
- EL: ER2 LockTense
- VL: R2 FixFrame (verb forms), keep time markers consistent

EL_SENSOR_MAP: EL.E3 LinkFailure

what_it_means: missing/wrong connectors; logic chain breaks
best_vl_islands: [VL.CI.logic, VL.CI.causality, VL.CI.coherence]
typical_vl_tripwires: [S1 SenseWrong (“therefore” misused)]
best_repairs:
- EL: ER3 InsertLinks (choose relation type)
- VL: R1 ReplaceByEdge (swap therefore/however/because), keep 1 mechanism line

EL_SENSOR_MAP: EL.E4 ReferenceLeak

what_it_means: pronouns unclear; missing nouns
best_vl_islands: [VL.CI.clarity, VL.CI.coherence]
typical_vl_tripwires: [S1 SenseWrong (vague nouns)]
best_repairs:
- EL: ER1 NameActors (repeat nouns)
- VL: R1 ReplaceByEdge (“her”→“Mei”; “thing”→specific noun)

EL_SENSOR_MAP: EL.E5 ScopeLeak

what_it_means: paragraph grows too big; new points drift in
best_vl_islands: [VL.CI.neutrality, VL.CI.coherence, VL.CI.uncertainty_hedging]
typical_vl_tripwires: [S5 IntensityMismatch; absolutes]
best_repairs:
- EL: ER4 ShrinkScope (1 arc / 2 reasons max)
- VL: R4 TuneIntensity + remove absolutes; hedge if needed

EL_SENSOR_MAP: EL.E6 LoadCollapse

what_it_means: run-ons, adjective stacking, compressed logic
best_vl_islands: [VL.CI.description, VL.CI.sensory, VL.CI.process]
typical_vl_tripwires: [S4 CollocationClash from stacking]
best_repairs:
- EL: ER5 SplitLoad (shorten; one idea per sentence)
- VL: R3 FixCollocation + cap sensory/description to 2 details

Mini “Use This Everywhere” Snippet (Universal Template)

Template:
1) Pick route_id
2) Draft EL skeleton only
3) Add 1–2 VL islands
4) Run sensors (EL then VL)
5) Apply repairs
6) Output v1

Recommended Internal Links (Spine)

Start Here for Lattice Infrastructure Connectors