@litfamily/litgrok 1.0.13 → 1.0.15

package/.grok/skills/lit-pptx/examples/11-business-review-night-en.md

@@ -0,0 +1,181 @@
+---
+tonality: night
+title: Parcel network operations review 2026, third quarter
+subtitle: Volumes, service level and cost for the operations town hall
+date: 2026-10-05
+department: Example Logistics, Network Operations
+presenter: Example Head of Operations
+notice: Sample data — replace with real figures
+---
+
+---
+layout: cover-numeral
+
+# Parcel network operations review 2026, third quarter
+---
+
+---
+layout: kpi-row
+
+## Q3 operating results against plan
+
+| Parcels per day | On-time delivery | Cost per parcel | Damage rate | Lost-time injuries |
+|---|---|---|---|---|
+| 412,000 | 96.1 % | 2.31 | 0.18 % | 4 |
+| plan 405,000; Q3 2025 371,000 | target 96.5 % | plan 2.36; Q3 2025 2.40 | target 0.20 % | Q3 2025: 7 |
+
+- Volume beat plan by 1.7 % and grew 11 % year on year
+- Cost per parcel came in 5 cents under plan after the east hub opened
+- On-time delivery missed target by 0.4 points, all of it in September
+- Source: sample network operations data mart, July-September 2025 and 2026 (sample)
+---
+
+---
+layout: full-chart
+
+## Parcels per day by month, 2025 and 2026
+
+::: chart type=column unit="thousand parcels"
+| Month | 2025 (k) | 2026 (k) |
+|---|---|---|
+| Apr | 352 | 389 |
+| May | 358 | 395 |
+| Jun | 361 | 398 |
+| Jul | 366 | 404 |
+| Aug | 369 | 409 |
+| Sep | 378 | 423 |
+> Average parcels handled per working day, April-September (sample data)
+:::
+
+- Source: sample network operations data mart, 2025-2026 (sample)
+---
+
+---
+layout: chart-insight
+
+## On-time delivery by region, Q3 2026
+
+::: chart type=bar unit="%"
+| Region | On-time (%) |
+|---|---|
+| North | 97.2 |
+| Capital | 96.8 |
+| East | 96.4 |
+| West | 95.9 |
+| South | 94.3 |
+> Share of parcels delivered by the promised day, Q3 2026 (sample data)
+:::
+
+- South is the only region below 95 %; the September storm closed two depots for three days
+- Without the storm days, South reached 96.0 % and the network 96.6 %
+- Capital improved 1.1 points after evening collection moved to the east hub
+- Source: sample delivery scan events, Q3 2026 (sample)
+---
+
+---
+layout: big-number
+
+## Cost per parcel and its two largest parts
+
+| Cost per parcel | Linehaul share | Last-mile share |
+|---|---|---|
+| 2.31 | 0.58 | 1.24 |
+| plan 2.36; Q3 2025 2.40 | Q3 2025 0.66 | Q3 2025 1.25 |
+
+- Linehaul cost fell 8 cents as trailers ran fuller through the east hub
+- Last-mile cost stayed flat; more stops per route offset higher driver pay
+- Fuel price added about 2 cents against Q3 2025
+- Source: sample management accounts, Q3 2025 and 2026 (sample)
+---
+
+---
+layout: section-numeral
+
+# Network changes
+---
+
+---
+layout: comparison
+
+## West hub and the new east hub, first full quarter
+
+:::: columns 1fr 1fr
+::: col
+- **West hub (2011)**
+  (1) Capacity: 180,000 parcels a day, run at 104 % in peak weeks
+  (2) Sort rate: 9,500 parcels an hour
+  (3) Linehaul fill: 71 % average trailer fill
+  (4) Mis-sort rate: 0.42 %
+:::
+::: col
+- **East hub (2026)**
+  (1) Capacity: 220,000 parcels a day, run at 78 % in Q3
+  (2) Sort rate: 16,000 parcels an hour
+  (3) Linehaul fill: 84 % average trailer fill
+  (4) Mis-sort rate: 0.11 %
+:::
+::::
+
+- Source: sample hub performance reports, Q3 2026 (sample)
+---
+
+---
+layout: dashboard-grid
+
+## Fleet use and fuel per 100 km
+
+::: chart type=line unit="%"
+| Month | Fleet utilisation (%) |
+|---|---|
+| Jul | 86 |
+| Aug | 88 |
+| Sep | 91 |
+> Delivery vans in service on an average working day (sample data)
+:::
+
+::: chart type=column unit="litres"
+| Month | Diesel per 100 km (L) |
+|---|---|
+| Jul | 11.8 |
+| Aug | 11.6 |
+| Sep | 11.9 |
+> Average diesel use per 100 km, delivery vans (sample data)
+:::
+
+- Utilisation above 90 % leaves no spare vans for the November peak without hires
+- Source: sample fleet telematics, July-September 2026 (sample)
+---
+
+---
+layout: timeline
+
+## Q4 plan to the November peak
+
+| Date | Milestone | Owner |
+|---|---|---|
+| 20 Oct | 60 hired vans on the road for peak | Fleet manager |
+| 3 Nov | South depot flood barriers installed | Facilities lead |
+| 10 Nov | Evening collection for West moves to the east hub | Hub manager, east |
+| 24 Nov | Peak week staffing at 118 % of normal | Regional directors |
+| 15 Dec | Peak review and 2027 hub plan | Head of operations |
+
+- Peak volume is forecast at 610,000 parcels a day in the week of 24 November
+- Source: sample peak plan, version 3 of 30 September 2026 (sample)
+---
+
+---
+layout: closing-decision-box
+
+## Decisions for the November peak
+
+- Decision requested: approve the van hires and the West evening move so peak service holds above 95 %
+
+| Decision | Cost | Date | Owner |
+|---|---|---|---|
+| Hire 60 vans for eight weeks | 0.72 M | 10 Oct | Fleet manager |
+| Move West evening collection to the east hub | none | 10 Oct | Head of operations |
+| Flood barriers for two South depots | 0.18 M | 17 Oct | Facilities lead |
+| Peak bonus for sort staff | 0.40 M | 17 Oct | HR director |
+
+- Next step: regional directors confirm peak staffing plans by 24 October
+---

package/.grok/skills/lit-pptx/examples/12-status-update-gazette-ko.md

@@ -0,0 +1,199 @@
+---
+tonality: gazette
+title: 차세대 회계 시스템 구축 9월 진행 보고
+subtitle: 진척, 예산, 위험과 10월 결정 사항
+date: 2026-10-05
+department: 예시 제조사 재무혁신 TF
+presenter: 예시 TF 팀장
+notice: 예시 데이터 — 실제 수치로 바꿔 주세요
+---
+
+---
+layout: cover-index
+
+# 차세대 회계 시스템 구축 9월 진행 보고
+---
+
+---
+layout: agenda
+
+## 보고 순서
+
+- **진행 요약** 9월 진척률과 일정 차이
+- **영역별 현황** 다섯 작업 영역의 진척과 예산
+- **이슈와 위험** 데이터 이관 지연과 대응
+- **결정 요청** 10월 경영회의 안건
+---
+
+---
+layout: summary-box-list
+
+## 9월 진행 요약
+
+::: key-message
+전체 진척률 58%로 계획보다 4%p 늦음. 데이터 이관 지연이 원인이며, 1월 가동 일정은 유지 가능
+:::
+
+- **진척**
+  (1) 계획 62% 대비 실적 58%, 8월 차이 −2%p에서 확대
+  (2) 다섯 영역 중 세 영역은 계획대로 진행
+  (3) 통합 테스트 1차는 10월 20일 시작 예정
+- **원인**
+  (1) 구 시스템 거래처 데이터 4만 2천 건 중 18%가 중복·누락
+  (2) 정제 작업에 현업 인력이 하루 6명 필요, 현재 3명 투입
+- **대응**
+  (1) 10월부터 정제 인력 3명 추가, 11월 7일까지 정제 완료
+  (2) 일정 여유 3주 중 2주를 사용해 1월 5일 가동은 유지
+- **예산**
+  (1) 9월 말 집행 16.4억 원, 총예산 28.0억 원의 59%
+  (2) 연말 전망 27.9억 원, 정제 인력 추가 0.4억 원은 예비비로 충당
+- **결정 요청**
+  (1) 10월 경영회의 안건 3건, 정제 인력 추가와 예비비 0.4억 원 전용
+  (2) 해외 미거래처 1,240건 이관 제외, 비용 없음
+---
+
+---
+layout: ledger-table
+
+## 작업 영역별 진척 현황
+
+| 작업 영역 | 계획 | 실적 | 차이 | 다음 산출물 | 담당 |
+|---|---|---|---|---|---|
+| 총계정원장 | 70% | 71% | +1%p | 결산 시나리오 10월 18일 | 회계팀 |
+| 채권·채무 | 65% | 63% | −2%p | 지급 승인 흐름 10월 25일 | 재무팀 |
+| 고정자산 | 60% | 60% | 0%p | 감가상각 검증 10월 30일 | 회계팀 |
+| 원가 | 55% | 54% | −1%p | 제품별 원가 배부 11월 8일 | 원가팀 |
+| 데이터 이관 | 58% | 41% | −17%p | 거래처 정제 완료 11월 7일 | 정보시스템팀 |
+| 전체 | 62% | 58% | −4%p | 통합 테스트 1차 10월 20일 | 재무혁신 TF |
+
+> 표 1. 작업 영역별 진척률, 2026년 9월 30일 기준 (예시 데이터)
+
+- 일정 차이의 대부분은 데이터 이관 영역에서 생김 (−17%p)
+- 나머지 네 영역은 계획 대비 ±2%p 안에 있음
+- 출처: 예시 제조사 재무혁신 TF 주간 진척 관리표, 2026년 9월 (예시)
+---
+
+---
+layout: table-insight
+
+## 예산 집행 현황
+
+| 항목 | 예산 | 집행 | 집행률 | 연말 전망 |
+|---|---|---|---|---|
+| 소프트웨어 사용료 | 9.6 | 7.2 | 75% | 9.6 |
+| 구축 용역 | 14.0 | 8.1 | 58% | 14.0 |
+| 데이터 정제 인력 | 1.2 | 0.7 | 58% | 1.6 |
+| 교육·변화 관리 | 1.5 | 0.4 | 27% | 1.4 |
+| 예비비 | 1.7 | 0.0 | 0% | 1.3 |
+| 합계 | 28.0 | 16.4 | 59% | 27.9 |
+
+> 표 2. 예산 집행, 단위 억 원, 2026년 9월 말 기준 (예시 데이터)
+
+- 정제 인력 추가 비용 0.4억 원은 예비비에서 충당
+- 연말 전망 27.9억 원으로 총예산 안에서 마무리 가능
+- 출처: 예시 제조사 재무혁신 TF 예산 관리표, 2026년 9월 (예시)
+---
+
+---
+layout: section-rule
+
+# 이슈와 위험
+---
+
+---
+layout: comparison
+title: side-rail
+
+## 거래처 데이터 정제, 9월과 10월 이후
+
+:::: columns 1fr 1fr
+::: col
+- **9월**
+  (1) 인력: 현업 3명, 필요 6명의 절반
+  (2) 처리량: 하루 약 700건
+  (3) 남은 물량: 2만 1천 건, 30일 분량
+  (4) 일정 여유: 3주
+:::
+::: col
+- **10월 6일 이후**
+  (1) 인력: 현업 6명, 필요 인원 충족
+  (2) 처리량: 하루 1,400건
+  (3) 남은 물량: 2만 1천 건, 15일 분량
+  (4) 일정 여유: 2주 사용, 1주 남음
+:::
+::::
+
+- 출처: 예시 제조사 정보시스템팀 데이터 이관 점검표, 2026년 9월 (예시)
+---
+
+---
+layout: text-column
+
+## 데이터 이관 지연 상세
+
+- **현황**
+  (1) 거래처 4만 2천 건 중 7,560건(18%)이 중복 또는 필수 항목 누락
+  (2) 정제 완료 2만 1천 건, 남은 2만 1천 건은 하루 약 700건씩 처리
+- **원인**
+  (1) 구 시스템에 사업장별로 같은 거래처가 따로 등록됨
+  (2) 사업자번호 없이 등록된 해외 거래처 1,240건
+- **대응**
+  (1) 10월 6일부터 정제 인력 3명 추가, 하루 처리량 1,400건으로 확대
+  (2) 해외 거래처는 구매팀이 거래 여부를 먼저 확인해 미거래처 제외
+- **일정 영향**
+  (1) 11월 7일 정제 완료 시 통합 테스트 2차는 11월 24일 시작 가능
+  (2) 일정 여유 3주 중 2주를 사용하고 1주가 남음
+- 출처: 예시 제조사 정보시스템팀 데이터 이관 점검표, 2026년 9월 (예시)
+---
+
+---
+layout: matrix-2x2
+
+## 주요 위험의 영향·가능성 배치
+
+| 영향 \ 가능성 | 높음 | 낮음 |
+|---|---|---|
+| 큼 | 정제 인력 확보 지연 · 1주 지연 · 인사팀 | 연말 결산과 테스트 일정 충돌 · 2주 지연 · 회계팀 |
+| 작음 | 사용자 교육 참석률 저조 · 가동 후 문의 증가 · TF | 인터페이스 사양 변경 · 3일 지연 · 정보시스템팀 |
+
+> 표 3. 주요 위험, 가능성은 앞으로 8주 기준 (예시)
+
+- 인력 확보는 10월 6일 투입이 확정되어 다음 보고에서 해소 예정
+- 결산 충돌을 피하려면 2차 테스트를 12월 첫 주 안에 끝내야 함
+- 영향이 작은 두 위험은 담당 팀이 주간 회의에서 따로 점검함
+- 출처: 예시 제조사 재무혁신 TF 위험 관리 대장, 2026년 9월 (예시)
+---
+
+---
+layout: timeline
+
+## 10-12월 추진 일정
+
+| 시점 | 할 일 | 담당 |
+|---|---|---|
+| 10월 20일 | 통합 테스트 1차 시작 (총계정원장·고정자산) | 회계팀 |
+| 11월 7일 | 거래처 데이터 정제 완료 | 정보시스템팀 |
+| 11월 24일 | 통합 테스트 2차 시작 (전 영역) | 재무혁신 TF |
+| 12월 5일 | 사용자 교육 완료, 380명 | 인사팀 |
+| 12월 19일 | 가동 전 최종 점검과 가동 승인 | 경영회의 |
+
+> 표 4. 추진 일정, 2026년 10-12월, 가동 목표 2027년 1월 5일 (예시)
+
+- 2차 테스트는 12월 첫 주까지 끝내 연말 결산 일정과 겹치지 않게 함
+- 거래처 정제가 11월 7일을 넘기면 2차 테스트와 가동 승인이 함께 밀림
+- 출처: 예시 제조사 재무혁신 TF 추진 계획, 2026년 9월 개정 (예시)
+---
+
+---
+layout: closing-ask
+
+## 10월 경영회의 결정 요청 사항
+
+| 안건 | 금액 | 기한 | 담당 |
+|---|---|---|---|
+| 정제 인력 3명 추가 투입 | 0.4억 원 | 10월 6일 | 인사팀 |
+| 예비비 0.4억 원 전용 | 0.4억 원 | 10월 10일 | 재무팀 |
+| 해외 미거래처 1,240건 이관 제외 | 비용 없음 | 10월 10일 | 구매팀 |
+
+- 다음 단계: 결정 후 TF 팀장이 10월 17일 주간 회의에서 수정 일정을 공유
+---

package/.grok/skills/lit-pptx/references/density-and-fill.md

@@ -0,0 +1,155 @@
+# Density and fill
+
+Two dials decide how a deck uses its pages. **Density** sets how much a page holds and how tightly:
+grid margins, type ramp, gaps, padding, table row height and how much empty band a slide may leave
+under its last block. **Variance** sets how many title treatments one deck mixes. Every pack starts at
+density 10; variance starts between 4 (gazette) and 8 (studio). The skill may move a dial by up to 2
+on its own and writes the move on the direction card; a larger move waits for the user. Set the dials
+in the frontmatter (`density: 8`) so the source records them.
+
+The engine cannot invent content, so a full page comes mostly from how the source is written. This
+file says what the dials change, what the engine does to close a gap, and how to write a source that
+fills.
+
+## Density steps
+
+| Density | Grid (16:9 side margin) | Ramp | Body | Gaps item / group / block | Padding | Min table row | Engine band cap |
+|---|---|---|---|---|---|---|---|
+| 1-2 | airy (60 pt) | presented | 18, steps up to 24 | 18 / 36 / 48 | 24 | 36 | 0.28 |
+| 3-4 | standard (48 pt) | presented | 18, steps up to 24 | 12 / 24 / 36 | 24 | 32 | 0.24 |
+| 5-6 | standard (48 pt) | reading | 14, steps up to 18 | 6 / 18 / 24 | 18 | 24 | 0.18 |
+| 7-8 | dense (36 pt) | reading | 14, steps up to 18 | 6 / 12 / 24 | 12 | 22 | 0.14 |
+| 9 | dense (36 pt) | reading | 14, no step-up | 6 / 12 / 24 | 12 | 20 | 0.12 |
+| 10 | compact (24 pt) | compact | 13, no step-up | 6 / 12 / 18 | 12 | 18 | 0.12 |
+
+The compact ramp is source 9, label 11, body 13, lead 18, title 26, display 36, cover 44 pt. On it a
+top-rule title sits close over its rule, the body starts about 92 pt from the top, a text column holds
+up to 22 lines, and type is never enlarged to fill. Nothing goes under the readable floor
+(body 12 pt, table cells and captions 11 pt, sources 9 pt). For a big room, density 8 gives body 14
+pt and 36 pt margins, and the skill may make that move without asking.
+
+## Variance
+
+| Variance | Most title treatments in one deck |
+|---|---|
+| 1-3 | 2 (only for a user who asks for a uniform deck) |
+| 4 | 3 |
+| 5-6 | 4 |
+| 7-10 | 5 |
+
+The cap never exceeds the number of treatments the pack has. At variance 4 a deck still needs five
+distinct layouts for the composition check, so variety has to come from the body partitions; see
+example 07, which pins two slides to reach them.
+
+## What the engine does with a gap
+
+When a content slide leaves more empty band than its cap, the engine tries the pack's `fill-order`
+one policy at a time and keeps a step only if nothing overflows:
+
+- **step-up**: body one ramp step larger, table data from label to body size, KPI values a step up.
+  Off on density 9 and 10.
+- **distribute**: item, group and block gaps open one step each; the body keeps its top edge, and a
+  row of peers keeps equal heights.
+- **anchor-visual**: a chart, figure or picture takes the remaining height; table rows grow up to a
+  quarter above the minimum.
+- **change-family**: an instruction to the author. The engine cannot pick a better shape for the
+  content, so the log says so.
+
+A column can stand empty beside a full one even when the band under the body is small: two short
+takeaways beside a chart, a source line at the column's foot and nothing between. The engine reads
+each text or picture column against the column beside it (its largest empty band, from the top of the
+pair down to where the longer column ends) and fills one that leaves more than 40 % of the body empty,
+by layout only and in this order:
+
+- the visual takes columns: a narrower takeaway column wraps onto more lines and a wider chart or
+  picture grows, down to three columns of takeaways, never lowering the takeaways' size;
+- the takeaways step up to lead size when they fit;
+- the room left is shared between the points, at most a sixth of the body a gap;
+- two short points beside a chart that still leave the column empty run under the chart, side by
+  side, and the chart takes the body's width;
+- a short sidebar note beside a long main column stands across the top at lead size and the main
+  points run in two columns under it;
+- a title under which a column still stays empty gives way to another title that fits.
+
+When none of these fills it, the gate fails the slide (OF-115) and the fix is in the source: the
+takeaway the slide lacks, the values beside the chart, or another family.
+
+Every attempt is logged:
+
+```text
+fill: slide 3 band 0.22 -> 0.06 (distribute)
+fill: slide 8 band 0.26 -> 0.26 (none: nothing on the slide can grow; choose a fuller family)
+```
+
+Title choice is part of filling: when the default title leaves a short body, the engine may switch to
+a rail or a bottom title so a chart or table gets the whole height. Padding cards and adding shapes are
+not in its repertoire.
+
+## Writing a source that fills
+
+These rules are what the twelve examples follow. Each gives the engine real material for the lower
+half of the body.
+
+- **Each content slide answers the reader's follow-up questions**: where a figure comes from, what it
+  is compared with (last year, the plan, a peer), which period it covers, who published it, and what
+  follows from it.
+- **Every data slide ends with a source line.** `출처: …` or `Source: …` names the origin and period;
+  a `주: …` or `Note: …` line may follow with a definition or caveat. Both go to the foot strip.
+- **Tables run five to eight rows**, four to six columns, units in the header or caption, two or
+  three takeaways after the caption. Three rows or fewer read better as a `kpi-row` or a
+  `comparison`.
+- **KPI rows hold four to six figures**, with a second row giving each figure's basis ("계획 대비
+  +2.8%", "Q2 1.71 M"). A `big-number` panel holds two to four related figures the same way.
+- **Comparisons have three or four criteria**, each opened with the same label on both sides
+  ("비용: …", "Accuracy: …"), so the engine can draw a criteria column and rows that read across.
+- **Charts carry a caption and two to four takeaways**; write the chart table with exact figures.
+- **Processes and step diagrams** name what happens at each step. Add two or three plain bullets
+  after the steps (rules, a sample sentence, the source); they run under the steps and fill the page.
+- **Method slides list eight to twelve terms** when the formula is short, including a worked value
+  for each term (example 04).
+- **Itemised text uses groups with sub-points**: four to six bold heads, two or three `(1)` `(2)`
+  sub-points each. Flat one-line bullets leave half a compact page empty.
+- **Timelines take five dated rows and two or three takeaways**; an axis with one line under it, or
+  nothing, can leave the band under the axis empty, which OF-109 fails.
+- **References list eight to ten entries**, sources first, then the definitions the deck relies on.
+- **Closings carry the ask, the decision, and the next step with owner and date**, usually as a
+  small table of item, amount or deadline, owner.
+- **Merge before padding.** A slide with little to say joins its neighbour or takes a family made
+  for short content. A bullet that restates the title is filler, not fill.
+
+## Numbers
+
+Figures come in structured groups, never as one giant number on its own page. Every figure has its
+unit, its basis or comparison, its period and its source. Nothing is set above title size. Sample
+values are marked: the frontmatter `notice:` line (`예시 데이터 — 실제 수치로 바꿔 주세요` or `Sample
+data — replace with real figures`) prints a notice on every slide, and captions and source lines say
+"(예시)" or "(sample)". Avoid inflated framing ("10배 성장", "massive"): state the before and after and
+let the reader judge.
+
+## How the gate measures fill
+
+On the compiled file, the empty band is the part of the body zone left under the last content block.
+The zone runs from the bottom of the title (or from the top margin under a bottom title) to the floor
+at 486 pt.
+Decoration, the footer and the notice are not content. Covers, sections, statements, quotes and
+statement closings are exempt.
+
+- **OF-112** lists slides above their density cap as advisories; the deck fails only when the median
+  band across measured slides exceeds 0.20.
+- **OF-109** is per slide and per card: one unbroken empty stretch over 35 % of the area fails, even
+  when the deck median is fine. Hollow text slides are what it usually catches.
+- **OF-111** fails when content slides share one composition (title zone × body partition × dominant
+  content) on more than 40 % of them, or show fewer distinct compositions than the deck length needs
+  (up to five).
+- **OF-115** fails a text or picture column whose largest empty band beside a longer column passes
+  40 % of the body; a source strip at its foot counts as a block but never hides the band above it,
+  and a chart's values table under the takeaways is read as part of their column. Beside the figure
+  rows of a kpi-row or big-number slide the takeaways' column is read down to the body floor, since
+  the rows stop where their spacing ends: the engine sets the takeaways across under the rows, or
+  lets them share the room beside the rows when the across form would leave a band between the two.
+- **Visual substance** fails a content slide whose blocks cover less than about a fifth of the
+  content area, and a deck in which every content slide is a table.
+
+When OF-109 fires on a text slide, add the missing supporting facts as groups and sub-points; when
+it fires on a comparison or timeline, add takeaways under it or pin a title whose body is wider.
+Shrinking type, stretching cards or adding shapes to pass the gate is a defect.