@litfamily/litcodex 1.0.15 → 1.0.17

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (123) hide show
  1. package/README-Ko-KR.md +10 -10
  2. package/README.md +10 -10
  3. package/dist/install/codex.d.ts +1 -9
  4. package/dist/install/codex.js +7 -1
  5. package/dist/install/generated-skill-payload-hashes.d.ts +101 -40
  6. package/dist/install/generated-skill-payload-hashes.js +101 -40
  7. package/dist/install/index.js +1 -0
  8. package/marketplace/plugins/litcodex/.codex-plugin/plugin.json +1 -1
  9. package/marketplace/plugins/litcodex/components/auto-update/package.json +1 -1
  10. package/marketplace/plugins/litcodex/components/comment-checker/package.json +1 -1
  11. package/marketplace/plugins/litcodex/components/git-bash/package.json +1 -1
  12. package/marketplace/plugins/litcodex/components/lit-loop/package.json +1 -1
  13. package/marketplace/plugins/litcodex/components/lsp/package.json +1 -1
  14. package/marketplace/plugins/litcodex/components/rules/package.json +1 -1
  15. package/marketplace/plugins/litcodex/components/start-work-continuation/package.json +1 -1
  16. package/marketplace/plugins/litcodex/components/telemetry/package.json +1 -1
  17. package/marketplace/plugins/litcodex/components/wikify-knowledge/package.json +1 -1
  18. package/marketplace/plugins/litcodex/skills/frontend-ui-ux/THIRD-PARTY-NOTICE.txt +0 -2
  19. package/marketplace/plugins/litcodex/skills/lit-docx/SKILL.md +16 -12
  20. package/marketplace/plugins/litcodex/skills/lit-docx/references/components.md +113 -0
  21. package/marketplace/plugins/litcodex/skills/lit-docx/references/direction-step.md +164 -0
  22. package/marketplace/plugins/litcodex/skills/lit-docx/references/examples/01-report-ko.md +95 -0
  23. package/marketplace/plugins/litcodex/skills/lit-docx/references/examples/02-brief-ko.md +45 -0
  24. package/marketplace/plugins/litcodex/skills/lit-docx/references/examples/03-manual-ko.md +105 -0
  25. package/marketplace/plugins/litcodex/skills/lit-docx/references/examples/04-proposal-ko.md +97 -0
  26. package/marketplace/plugins/litcodex/skills/lit-docx/references/examples/05-memo-en.md +34 -0
  27. package/marketplace/plugins/litcodex/skills/lit-docx/references/examples/06-journal-en.md +89 -0
  28. package/marketplace/plugins/litcodex/skills/lit-docx/references/examples/07-report-en.md +101 -0
  29. package/marketplace/plugins/litcodex/skills/lit-docx/references/examples/08-proposal-en.md +94 -0
  30. package/marketplace/plugins/litcodex/skills/lit-docx/references/examples/09-memo-ko.md +34 -0
  31. package/marketplace/plugins/litcodex/skills/lit-docx/references/page-composition.md +141 -0
  32. package/marketplace/plugins/litcodex/skills/lit-docx/references/tonalities/brief.md +70 -0
  33. package/marketplace/plugins/litcodex/skills/lit-docx/references/tonalities/journal.md +73 -0
  34. package/marketplace/plugins/litcodex/skills/lit-docx/references/tonalities/manual.md +70 -0
  35. package/marketplace/plugins/litcodex/skills/lit-docx/references/tonalities/memo.md +75 -0
  36. package/marketplace/plugins/litcodex/skills/lit-docx/references/tonalities/proposal.md +71 -0
  37. package/marketplace/plugins/litcodex/skills/lit-docx/references/tonalities/report.md +83 -0
  38. package/marketplace/plugins/litcodex/skills/lit-docx/scripts/convert_md_to_docx.py +401 -58
  39. package/marketplace/plugins/litcodex/skills/lit-docx/scripts/docx_craft.py +124 -1
  40. package/marketplace/plugins/litcodex/skills/lit-docx/scripts/docx_design.py +2446 -0
  41. package/marketplace/plugins/litcodex/skills/lit-docx/scripts/docx_layout.py +960 -0
  42. package/marketplace/plugins/litcodex/skills/lit-docx/scripts/slop_lint.py +21 -6
  43. package/marketplace/plugins/litcodex/skills/lit-docx/templates/tonalities/brief.yaml +28 -0
  44. package/marketplace/plugins/litcodex/skills/lit-docx/templates/tonalities/journal.yaml +26 -0
  45. package/marketplace/plugins/litcodex/skills/lit-docx/templates/tonalities/manual.yaml +26 -0
  46. package/marketplace/plugins/litcodex/skills/lit-docx/templates/tonalities/memo.yaml +27 -0
  47. package/marketplace/plugins/litcodex/skills/lit-docx/templates/tonalities/proposal.yaml +27 -0
  48. package/marketplace/plugins/litcodex/skills/lit-docx/templates/tonalities/report.yaml +27 -0
  49. package/marketplace/plugins/litcodex/skills/lit-pptx/SKILL.md +12 -11
  50. package/marketplace/plugins/litcodex/skills/lit-pptx/pretendard-font/public/static/Pretendard-Bold.otf +0 -0
  51. package/marketplace/plugins/litcodex/skills/lit-pptx/pretendard-font/public/static/Pretendard-Regular.otf +0 -0
  52. package/marketplace/plugins/litcodex/skills/lit-pptx/references/anti-slop-checklist.md +33 -29
  53. package/marketplace/plugins/litcodex/skills/lit-pptx/references/authoring-guide.md +11 -2
  54. package/marketplace/plugins/litcodex/skills/lit-pptx/references/density-and-fill.md +179 -0
  55. package/marketplace/plugins/litcodex/skills/lit-pptx/references/design-system-azure.md +1 -1
  56. package/marketplace/plugins/litcodex/skills/lit-pptx/references/direction-step.md +203 -0
  57. package/marketplace/plugins/litcodex/skills/lit-pptx/references/examples/01-pitch-signal-ko.md +185 -0
  58. package/marketplace/plugins/litcodex/skills/lit-pptx/references/examples/02-pitch-studio-en.md +162 -0
  59. package/marketplace/plugins/litcodex/skills/lit-pptx/references/examples/03-business-review-ledger-ko.md +229 -0
  60. package/marketplace/plugins/litcodex/skills/lit-pptx/references/examples/04-research-talk-paper-en.md +212 -0
  61. package/marketplace/plugins/litcodex/skills/lit-pptx/references/examples/05-research-keynote-night-ko.md +191 -0
  62. package/marketplace/plugins/litcodex/skills/lit-pptx/references/examples/06-lecture-chalk-ko.md +175 -0
  63. package/marketplace/plugins/litcodex/skills/lit-pptx/references/examples/07-data-review-night-en.md +241 -0
  64. package/marketplace/plugins/litcodex/skills/lit-pptx/references/examples/08-field-report-atlas-ko.md +168 -0
  65. package/marketplace/plugins/litcodex/skills/lit-pptx/references/examples/09-briefing-gazette-ko.md +219 -0
  66. package/marketplace/plugins/litcodex/skills/lit-pptx/references/examples/10-status-update-ledger-en.md +157 -0
  67. package/marketplace/plugins/litcodex/skills/lit-pptx/references/examples/11-portfolio-studio-ko.md +183 -0
  68. package/marketplace/plugins/litcodex/skills/lit-pptx/references/examples/12-tutorial-chalk-en.md +176 -0
  69. package/marketplace/plugins/litcodex/skills/lit-pptx/references/examples/13-status-update-gazette-ko.md +157 -0
  70. package/marketplace/plugins/litcodex/skills/lit-pptx/references/interview-protocol.md +1 -1
  71. package/marketplace/plugins/litcodex/skills/lit-pptx/references/layout-families.md +210 -0
  72. package/marketplace/plugins/litcodex/skills/lit-pptx/references/title-treatments.md +146 -0
  73. package/marketplace/plugins/litcodex/skills/lit-pptx/references/tonalities/atlas.md +113 -0
  74. package/marketplace/plugins/litcodex/skills/lit-pptx/references/tonalities/chalk.md +115 -0
  75. package/marketplace/plugins/litcodex/skills/lit-pptx/references/tonalities/gazette.md +118 -0
  76. package/marketplace/plugins/litcodex/skills/lit-pptx/references/tonalities/ledger.md +129 -0
  77. package/marketplace/plugins/litcodex/skills/lit-pptx/references/tonalities/night.md +126 -0
  78. package/marketplace/plugins/litcodex/skills/lit-pptx/references/tonalities/paper.md +109 -0
  79. package/marketplace/plugins/litcodex/skills/lit-pptx/references/tonalities/signal.md +121 -0
  80. package/marketplace/plugins/litcodex/skills/lit-pptx/references/tonalities/studio.md +111 -0
  81. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/compile-deck.js +65 -17
  82. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/craft_extras.py +654 -11
  83. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/embed_fonts.py +1 -1
  84. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/inventory.py +32 -2
  85. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/lib/directive-parser.js +2 -2
  86. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/lib/grid-resolver.js +521 -0
  87. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/lib/layout-resolver.js +401 -16
  88. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/lib/markdown-loader.js +6 -4
  89. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/lib/render-adapter-html.js +2 -2
  90. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/lib/render-adapter-pptx.js +498 -145
  91. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/lib/render-pack.js +3558 -0
  92. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/lib/slide-ast.js +54 -6
  93. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/lib/template-registry.js +335 -11
  94. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/office-run.mjs +2 -1
  95. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/qa_deck.py +99 -19
  96. package/marketplace/plugins/litcodex/skills/lit-pptx/scripts/validate_pptx.py +6 -1
  97. package/marketplace/plugins/litcodex/skills/lit-pptx/templates/boilerplate_professional.css +6 -6
  98. package/marketplace/plugins/litcodex/skills/lit-pptx/templates/enrolled/AZURE-A2Z/layout-mapping.yaml +1 -1
  99. package/marketplace/plugins/litcodex/skills/lit-pptx/templates/enrolled/AZURE-A2Z/template.yaml +4 -1
  100. package/marketplace/plugins/litcodex/skills/lit-pptx/templates/enrolled/AZURE-PRO/layout-mapping.yaml +1 -1
  101. package/marketplace/plugins/litcodex/skills/lit-pptx/templates/enrolled/AZURE-PRO/template.yaml +4 -1
  102. package/marketplace/plugins/litcodex/skills/lit-pptx/templates/tonalities/atlas/pack.yaml +105 -0
  103. package/marketplace/plugins/litcodex/skills/lit-pptx/templates/tonalities/chalk/pack.yaml +110 -0
  104. package/marketplace/plugins/litcodex/skills/lit-pptx/templates/tonalities/gazette/pack.yaml +105 -0
  105. package/marketplace/plugins/litcodex/skills/lit-pptx/templates/tonalities/ledger/pack.yaml +106 -0
  106. package/marketplace/plugins/litcodex/skills/lit-pptx/templates/tonalities/night/pack.yaml +106 -0
  107. package/marketplace/plugins/litcodex/skills/lit-pptx/templates/tonalities/paper/pack.yaml +111 -0
  108. package/marketplace/plugins/litcodex/skills/lit-pptx/templates/tonalities/signal/pack.yaml +107 -0
  109. package/marketplace/plugins/litcodex/skills/lit-pptx/templates/tonalities/studio/pack.yaml +105 -0
  110. package/marketplace/plugins/litcodex/skills/lit-typographic-motion/engine/THIRD_PARTY_NOTICES +1 -1
  111. package/marketplace/plugins/litcodex/skills/lit-typographic-motion/engine/fonts.mjs +6 -6
  112. package/marketplace/plugins/litcodex/skills/lit-typographic-motion/engine/gate.mjs +1 -1
  113. package/marketplace/plugins/litcodex/skills/lit-typographic-motion/engine/stage-qa.mjs +2 -2
  114. package/marketplace/plugins/litcodex/skills/lit-typographic-motion/references/runtime.md +1 -1
  115. package/marketplace/plugins/litcodex/skills/lit-typographic-motion/references/stage.md +1 -1
  116. package/marketplace/plugins/litcodex/skills/lit-typographic-motion/references/style-bibles.md +2 -2
  117. package/marketplace/plugins/litcodex/skills/lit-typographic-motion/references/treatment.md +1 -1
  118. package/marketplace/plugins/litcodex/skills/lit-typographic-motion/references/type-and-timing.md +3 -3
  119. package/node_modules/@litcodex/lit-loop/package.json +1 -1
  120. package/package.json +2 -2
  121. package/readme-assets/badge-version.svg +1 -1
  122. package/marketplace/plugins/litcodex/skills/lit-pptx/pretendard-font/public/static/PretendardGOV-Bold.otf +0 -0
  123. package/marketplace/plugins/litcodex/skills/lit-pptx/pretendard-font/public/static/PretendardGOV-Regular.otf +0 -0
@@ -0,0 +1,179 @@
1
+ # Density and fill
2
+
3
+ Two dials govern how a deck uses its pages. **Density** sets how much a page holds and how tightly
4
+ it is set. **Variance** sets how many title treatments and families one deck mixes. Every pack
5
+ ships with defaults (density 10 in all eight; variance 4-8), and the direction card records the
6
+ values actually used. Moves of 1 or 2 on either dial are the skill's call, noted on the card;
7
+ anything further waits for the user.
8
+
9
+ Fill is the reason density exists. Earlier decks put a short table at the top of a slide, stretched
10
+ cards to the height of their two lines, ended on a bare thank-you, and left the lower half of most
11
+ pages empty. The compiler now measures the empty band under each slide's last block, tries to
12
+ close it with the pack's fill policies, and the gate fails a deck whose median slide still wastes
13
+ more than a fifth of its body. The compiler can only grow what the source gives it, so most fill
14
+ is won or lost in how the source is written.
15
+
16
+ ## The density dial
17
+
18
+ | Density | Grid, 16:9 side margin | Ramp | Body | Gaps item / group / block | Padding | Min table row | Empty band allowed |
19
+ |---|---|---|---|---|---|---|---|
20
+ | 1-2 | airy, 60 pt | presented | 24 for four items or fewer, else 18 | 18 / 36 / 48 | 24 | 36 | 28 % |
21
+ | 3-4 | standard, 48 pt | presented | 18, may step to 24 | 12 / 24 / 36 | 24 | 32 | 24 % |
22
+ | 5-6 | standard, 48 pt | reading | 14, may step to 18 | 6 / 18 / 24 | 18 | 24 | 18 % |
23
+ | 7-8 | dense, 36 pt | reading | 14, may step to 18 | 6 / 12 / 24 | 12 | 22 | 14 % |
24
+ | 9 | dense, 36 pt | reading | 14 | 6 / 12 / 24 | 12 | 20 | 12 % |
25
+ | **10 (default)** | compact, 24 pt | compact | 13 | 6 / 12 / 18 | 12 | 18 | 12 % |
26
+
27
+ **The compact step.** Density 10 uses the compact grid (24 pt margins, 54 pt columns, 24 pt
28
+ gutters, 78 pt pitch on 16:9) and the compact ramp: source 9, label 11, body 13, lead 18, title 26,
29
+ display 36, cover 44. A top-rule title sits tight over its rule, so a one-line title lets the body
30
+ start near y 92 instead of 120. Up to 22 lines fit a text column. Type is never enlarged to fill on
31
+ this step. The readable floor holds everywhere: body 12 pt or more, table cells and captions 11 pt
32
+ or more, sources 9 pt or more.
33
+
34
+ **When to lower it.** Drop to 8 (dense grid, body 14 pt) for a deck shown on a stage or to a large
35
+ room and not read afterwards. Going below 8 needs the user's word. Never raise legibility by
36
+ shrinking content instead: cut or split the slide.
37
+
38
+ ## The variance dial
39
+
40
+ | Variance | Treatments per deck | Same family back to back |
41
+ |---|---|---|
42
+ | 1-3 (only on request) | 1-2; the gate records "variety lowered by the user" | — |
43
+ | 4 | 3 | up to 3 |
44
+ | 5-6 | 3-4 | up to 3 |
45
+ | 7-8 | 4-5 | up to 2 |
46
+ | 9-10 | 5 | up to 2, and no family more than twice in the deck |
47
+
48
+ The ceiling is also bounded by how many treatments the pack owns (four in Chalk, Gazette, Ledger,
49
+ Paper and Signal; five in Atlas, Night and Studio).
50
+
51
+ ## Fill policies
52
+
53
+ When a content slide's empty band exceeds its density's allowance, the compiler walks the pack's
54
+ `fill-order` and keeps a step only if nothing overflows.
55
+
56
+ | Policy | What it does | Limits |
57
+ |---|---|---|
58
+ | `step-up` | body one ramp step larger, table data from label to body size, KPI values one step up | one step; never on the compact step; titles never shrink to make room |
59
+ | `distribute` | opens item, group and block gaps one step each, top edge fixed | peers keep equal heights; the body is never centred vertically |
60
+ | `anchor-visual` | lets a chart, figure or picture take the remaining height; table rows grow up to a quarter over the minimum | only when the slide has a visual |
61
+ | `change-family` | none by itself: the log says "choose a fuller family" | the author changes the source |
62
+
63
+ | Pack | fill-order |
64
+ |---|---|
65
+ | Atlas | anchor-visual, change-family, distribute, step-up |
66
+ | Chalk | distribute, step-up, change-family, anchor-visual |
67
+ | Gazette | step-up, distribute, change-family, anchor-visual |
68
+ | Ledger | step-up, distribute, anchor-visual, change-family |
69
+ | Night | anchor-visual, step-up, distribute, change-family |
70
+ | Paper | anchor-visual, distribute, step-up, change-family |
71
+ | Signal | step-up, change-family, anchor-visual, distribute |
72
+ | Studio | change-family, step-up, anchor-visual, distribute |
73
+
74
+ Each filled or unfillable slide leaves a line in the compile output, which belongs in the build log:
75
+
76
+ ```text
77
+ fill: slide 6 band 0.38 -> 0.11 (distribute)
78
+ fill: slide 11 band 0.33 -> 0.33 (none: nothing on the slide can grow; choose a fuller family)
79
+ ```
80
+
81
+ A column can stand empty beside a full one even when the body's last block reaches the floor: two
82
+ short takeaways beside a chart, a source line at the column's foot and nothing between. The engine
83
+ reads each text or picture column against the column beside it (its largest empty band, from the top
84
+ of the pair down to where the longer column ends) and fills a column that leaves more than 40 % of the
85
+ body empty, by layout only and in this order. When the compact step sets a chart's values as a small
86
+ table under the takeaways, that table belongs to their column: it moves down with them and takes a
87
+ share of the room like a point does.
88
+
89
+ - the visual takes columns: a narrower takeaway column wraps onto more lines while a wider chart or
90
+ picture grows (down to three columns of takeaways; a picture keeps taking columns while either side
91
+ leaves a fifth of the body empty), and no takeaway loses size for it;
92
+ - the takeaways step up to lead size when they fit and keep twelve glyphs a line;
93
+ - the room left is shared between the points, never more than a sixth of the body in one gap;
94
+ - two short points that still leave the column empty beside a chart stand under it side by side, and
95
+ the chart with its values takes the body's width;
96
+ - a short sidebar note beside a long main column runs across the top at lead size, with the main
97
+ points in two columns under it;
98
+ - summary groups of unequal length, split by count, can leave one column far shorter than the other;
99
+ the split then moves to the group boundary that evens the columns most, and a head never leaves its
100
+ points;
101
+ - a title under which a column still stays empty is unfit for the slide when another title fits.
102
+
103
+ The source strip at a column's foot counts as a block but never hides the band above it, and the
104
+ gate measures the same way (OF-115), with two additions: it reads a takeaway column that holds the
105
+ chart's values table (at the table's drawn height), and it reads every pair of columns whatever the
106
+ longer one's length, so a short note beside points that stop at two thirds of the body still fails.
107
+ Beside the figure rows of a kpi-row or big-number slide both read the takeaways' column down to the
108
+ body floor, since the rows stop where their spacing ends: short points run under the rows, or share
109
+ the room beside them when running under would leave a band between rows and points.
110
+ What the engine cannot refill, the source fixes: two or three takeaways computed from the slide's own
111
+ figures, or a real further group where the slide's subject has one; never smaller type or a decoration.
112
+
113
+ The treatment choice is also a fill decision: a slide whose body stays short under its role default
114
+ may be drawn under a side rail or a bottom title, where the visual gets the full height. No policy
115
+ adds decoration or pads a card to cover space.
116
+
117
+ ## Writing a source that fills
118
+
119
+ Every rule below hands the compiler something real to place lower on the page.
120
+
121
+ - **Five supporting facts per content slide.** Beside the main point, give what a careful reader
122
+ checks: the basis of each figure, a comparison (prior period, plan, peer), the period, the source,
123
+ and the implication or next action.
124
+ - **End data slides with a strip.** A line `출처: …` / `Source: …` naming origin and date; optionally
125
+ `주: …` / `Note: …` with a definition or caveat. These cost the takeaways no room, and the compiler
126
+ moves them to the foot wherever they stand in the slide text. The examples keep a Korean `출처:`
127
+ line directly under the table or picture it describes, so the attribution sits next to its
128
+ evidence in the source file. On slides without a table or picture, and on the first slides after
129
+ a section (whose opening lines a section page may preview), they fold the source into the `주:`
130
+ line instead ("주: … 자료는 …").
131
+ - **KPI rows of four to six.** The table's second body row is each figure's basis ("목표 대비 +2.1%p",
132
+ "전년 동기 1,940건"). Without it the row reads as decoration and the slide under-fills. Keep each
133
+ value to 12 characters and each basis cell to 24; longer cells stop the table being read as a KPI
134
+ row, and `cover-figures` then finds no headline figures and falls back to the plain cover.
135
+ - **Tables with real depth:** five to eight rows, four to six columns when the source has that much,
136
+ units in the headers, totals in bold. Three rows or fewer usually wants `kpi-row` or `comparison` instead.
137
+ - **Two or three takeaways under every table and chart.** The row that matters, the comparison, the
138
+ decision it supports. The chart's caption says what is plotted and marks the data as example data
139
+ when it is.
140
+ - **Exact chart tables.** On the compact step a chart's values (up to eight rows, four columns) are
141
+ shown as a small table under its takeaways, so write them with real precision and units.
142
+ - **Labelled comparison points.** Same label on both sides ("비용: …", "일정: …"), three or four
143
+ criteria, each a full clause, equal count per side; they draw as a criteria column with rows that
144
+ reach the floor together.
145
+ - **Steps that say what happens.** A process or timeline row names the action, the owner and the
146
+ measure, not only the step name. A timeline's takeaways stand under its axis, so give it two or three
147
+ of them: a single line leaves the band under the axis empty, and OF-109 fails it.
148
+ - **Panels, not lone numbers.** A `big-number` holds two to four related figures with labels and a
149
+ basis row, a caption with period and source, and two or three lines of evidence.
150
+ - **A closing that asks.** The title labels the request; the body tables what is to be approved (item,
151
+ amount, date, owner) and names the next step with its owner and date.
152
+ - **References with dates,** one per row, eight to ten entries; with only a few sources, add the
153
+ definitions the deck relies on.
154
+ - **The label is the title, the claim is the body.** "권역별 누수율 추이", not "누수율이 세 분기 연속
155
+ 줄었다"; the sentence becomes the first takeaway.
156
+ - **No inflated numbers.** Report figures at the precision the source supports, keep units, and do
157
+ not promote a modest change into a headline with words like "폭증" or "game-changing". A figure
158
+ that appears twice is identical both times.
159
+ - **Merge before padding.** If a slide still has little to say, merge it into its neighbour or pick a
160
+ family built for short content. Filler bullets that restate the title are a defect.
161
+
162
+ ## How the gate measures fill
163
+
164
+ `qa_deck.py` works on the compiled file. The empty band is the part of the body under its lowest
165
+ content block, as a share of the body height; the body runs from the title's lower edge (or from the
166
+ top margin when the title sits at the bottom) to y 486. Decoration, the footer and the example notice do not count as content. Covers, sections,
167
+ statements, quotes and statement closings are exempt.
168
+
169
+ | Check | Rule |
170
+ |---|---|
171
+ | OF-112 | per slide: an advisory with a fix hint when the band exceeds the density allowance; per deck: fail when the median band over measured slides is above 0.20 |
172
+ | OF-109 | fail when any single slide or card leaves more than 35 % of its area as one empty band |
173
+ | OF-115 | fail when a region stands empty, among them a text or picture column whose largest empty band beside a longer column passes 40 % of the body (a source strip at its foot included) |
174
+ | OF-111 | fail when one composition (title zone × body partition × dominant content) covers more than 40 % of content slides, or too few distinct compositions appear |
175
+
176
+ A slide that cannot be filled honestly (a five-line agenda, a two-source appendix) may stay over its
177
+ allowance; the deck median must still hold. Shrinking text, inflating cards or adding shapes to pass
178
+ the gate is itself a defect that the critique records. On a legacy template the median check only
179
+ advises, since the legacy renderer cannot apply fill policies.
@@ -1,6 +1,6 @@
1
1
  # Azure Professional — SOTA Blue & White Design System (2026)
2
2
 
3
- A pro-designer blue-and-white system for `BOILERPLATE-AZURE` (flagship). 16:9 widescreen.
3
+ A blue-and-white system for the legacy `AZURE-PRO` template, 16:9 widescreen. It applies only when the user names that template; new decks take a tonality pack from the direction card.
4
4
  Goal aesthetic: modern, geometric, confident — like a paid premium template, not a stock deck.
5
5
 
6
6
  ## Color tokens (locked)
@@ -0,0 +1,203 @@
1
+ # Direction step
2
+
3
+ A deck's look is decided once, at the start, and written down. The decision names a tonality: a
4
+ pack of palette, faces, title treatments, layout families, decoration and two dials (density,
5
+ variance) that the compiler reads as data. Eight tonalities ship with this skill; four older
6
+ single-look templates stay available for users who ask for them by name. Skipping the step is
7
+ how every deck ends up blue and white regardless of subject, so the build does not start until
8
+ the direction card below exists.
9
+
10
+ The step costs a few minutes. Close the brief, take the deck type's candidates, count the outline,
11
+ reorder, write the card, and tell the user the pick with two alternatives. Questions to the user
12
+ are reserved for two cases: the brief leaves the look open on purpose (build the compare strip),
13
+ or a dial needs a bigger move than the skill is allowed to make unasked.
14
+
15
+ ## Inputs: four facts about the brief
16
+
17
+ Write these down before ranking. Where the brief is silent, infer from the request and the
18
+ attached material, and mark the line `(inferred)` so the user can overrule it.
19
+
20
+ | Fact | What to record | Why it matters |
21
+ |---|---|---|
22
+ | Deck type | one of the eight rows below, or "other, closest to …" | picks the candidate list |
23
+ | Reader | who, in what setting, for how long ("투자 심사역 4명, 회의실, 15분") | presented vs read, glyph budget |
24
+ | Delivery | spoken live, sent ahead and read alone, or both | moves Gazette and Ledger |
25
+ | Language | Korean, English, mixed | Korean bodies carry about half the characters of Latin ones |
26
+
27
+ ## Candidates by deck type
28
+
29
+ Each type has an ordered list. The first entry wins when no signal moves the order. Every pack
30
+ now starts at density 10 (the compact step); the variance shown is the pack default.
31
+
32
+ | Deck type | First | Second | Third |
33
+ |---|---|---|---|
34
+ | 투자·제안 피치 (pitch) | Signal, v6 | Studio, v8 | Atlas, v7 when the product is something people look at |
35
+ | 경영 실적 보고 (business review) | Ledger, v5 | Gazette, v4 when circulated in Korean and read alone | Night, v6 when shown on a stage screen |
36
+ | 연구 발표 (research talk) | Paper, v5 | Night, v6 for a keynote hall | Chalk, v6 for a tutorial-style talk |
37
+ | 강의 (lecture, teaching) | Chalk, v6 | Paper, v5 | Studio, v8 for a public lecture |
38
+ | 데이터 리뷰 (data-heavy review) | Night, v6 | Ledger, v5 | Paper, v5 for a written analysis |
39
+ | 이미지 중심 (product, site, portfolio) | Atlas, v7 | Studio, v8 | Signal, v6 for a launch with few pictures |
40
+ | 보고서형 덱 (Korean text briefing) | Gazette, v4 | Ledger, v5 | — |
41
+ | 주간·월간 보고 (status update) | Ledger, v5 | Gazette, v4 | Night, v6 |
42
+
43
+ ## Reordering signals
44
+
45
+ Draft the outline first: one line per slide with its message and the visual it needs (table,
46
+ chart, picture, steps, quote). Count on that outline, never on the raw source. Apply the rows in
47
+ this order; a tie keeps the deck-type order, which makes the result repeatable.
48
+
49
+ | # | Signal | Fires when | Effect |
50
+ |---|---|---|---|
51
+ | 1 | Named tonality | the brief or an earlier turn names one | use it; the card is still written |
52
+ | 2 | Named legacy template | AZURE-PRO, AZURE-A2Z, BOILERPLATE-PRETENDARD or BOILERPLATE-A2Z is asked for | use it as named; offer two tonalities as alternatives |
53
+ | 3 | Read, not presented | "미리 배포", "보고서", "circulated", "pre-read" | Gazette (Korean) or Ledger (other languages) to the front |
54
+ | 4 | Table share | 30 % of content slides or more carry a table | Ledger or Gazette up one |
55
+ | 5 | Chart share | 30 % or more carry a chart | Ledger, Night or Paper up one |
56
+ | 6 | Picture share | 30 % or more carry a photo, render or screenshot | Atlas or Studio up one; Gazette down one |
57
+ | 7 | Long bodies | average 150 Korean glyphs (300 Latin characters) or more per content slide | Gazette or Ledger up; Signal down |
58
+ | 8 | Short bodies | average 40 Korean glyphs (80 Latin characters) or fewer | Signal or Studio up; density stays 10, because source strips and basis rows fill the room |
59
+ | 9 | Citations | numbered references, DOIs or a references slide | Paper up one |
60
+ | 10 | Teaching structure | definitions, formulas, or a how-it-works sequence of three steps or more | Chalk or Paper up one |
61
+ | 11 | Dark room | keynote, stage, demo day | Night up one |
62
+ | 12 | Missing shape | the pack has no family or treatment for one of the outline's slides | drop that pack |
63
+
64
+ Row 12 decides more often than it seems. Atlas carries no table family. Ledger and Gazette have no
65
+ `statement` treatment or family. Signal has no `ledger-table` and no `dashboard-grid`. Paper has no
66
+ `kpi-row`. Chalk has no chart family. Gazette has no picture family. Check with `--list-layouts`
67
+ before committing:
68
+
69
+ ```bash
70
+ node "$OFFICE" pptx compile --list-tonalities
71
+ node "$OFFICE" pptx compile --list-layouts night
72
+ ```
73
+
74
+ **Worked case.** A twelve-slide English review of a city water network's leak programme, for the
75
+ utility board and its analysts, presented in the board room and filed afterwards. Type: data-heavy
76
+ review, so Night, Ledger, Paper. Of ten content slides, five hold a chart (50 %) and two a table
77
+ (20 %). Row 5 fires and lifts all three candidates alike, so it leaves their order unchanged; row 4
78
+ stays below its threshold. Bodies average about 210 Latin characters, so neither text row fires,
79
+ and nothing says the file is sent ahead. Night stays first. Had a third table slide been planned,
80
+ row 4 would have lifted Ledger past Night; the card says so in its "wrong if" line, together with
81
+ the printed board pack, which favours Ledger's white pages.
82
+
83
+ ## The direction card
84
+
85
+ The card is the decision on record. Every field is filled; a build that cannot write the tonality
86
+ and its reason stops and asks, because there is no silent default look.
87
+
88
+ | Field | Content |
89
+ |---|---|
90
+ | Deck type | the type row used, or "other, closest to …" |
91
+ | Reader | who, where, how long |
92
+ | Tonality | the chosen pack id, then the two alternatives |
93
+ | Reason | one sentence linking the pick to the reader and to the signals that fired |
94
+ | Signals fired | the row numbers or names from the table above |
95
+ | Dials | density and variance, with any move from the pack default and its reason |
96
+ | Treatments | the role → treatment map from the tonality sheet, with any planned pins |
97
+ | Families planned | one family per slide, in deck order, each from the pack's list |
98
+ | Wrong if | the one fact that, if true, means the other alternative should have won |
99
+
100
+ ```text
101
+ Direction card
102
+ Deck type data-heavy review (데이터 리뷰)
103
+ Reader utility board and two analysts, board room projector, 20 minutes; filed afterwards
104
+ Tonality night alternatives: ledger, paper
105
+ Reason chart-led review watched on a projector: 50 % chart slides, tables under 30 %
106
+ Signals fired 5 chart share
107
+ Dials density 10, variance 6 (pack defaults)
108
+ Treatments content top-plain-large · data band · data-takeaway side-rail · sequence kicker-numeral
109
+ Families planned cover-figures, kpi-row, full-chart, chart-insight, chart-insight, dashboard-grid,
110
+ section-numeral, big-number, timeline, comparison, kpi-over-chart, closing-decision-box
111
+ Wrong if the deck is printed into the board pack, or a third table slide joins; then Ledger
112
+ ```
113
+
114
+ **Dials.** Density (1-10) sets margins, gaps, body size and the empty band a slide may leave.
115
+ Variance (1-10) caps how many title treatments and families one deck mixes and how often a family
116
+ may repeat. Moves of 1 or 2 on either dial are the skill's own call and go on the card; bigger
117
+ moves wait for the user. The usual move is density 10 → 8 for a deck shown to a
118
+ large hall. Variance under 4 happens only when the user wants a uniform deck; the gate then notes
119
+ the reduced variety rather than passing it quietly. See `density-and-fill.md`.
120
+
121
+ **Treatments and families.** Copy the role map from the tonality sheet; pick a family per slide
122
+ from the pack's list. The compiler may still choose another allowed treatment for a slide whose
123
+ body fills better under it, and the compile log names each such move.
124
+
125
+ ## Telling the user, keeping the log
126
+
127
+ The reply states the pick, the reason in plain words, and the two alternatives, in two or three
128
+ sentences, before or with the files. For example: "Night로 만들었습니다. 차트가 슬라이드 절반이고
129
+ 이사회가 프로젝터로 보는 자료라서입니다. 인쇄해서 배포한다면 Ledger, 서면 분석으로 남긴다면 Paper가
130
+ 더 맞습니다."
131
+
132
+ Keep a build log next to the source. It holds the card, the compile output's `note:`, `fill:`,
133
+ `variety:` and `Treatments:` lines, and the gate result. It is working evidence; it is not a slide
134
+ and is handed over only when the user asks for the audit trail.
135
+
136
+ ```text
137
+ <deck>.md source with tonality: in the frontmatter
138
+ <deck>.build.md direction card, compile lines, gate result
139
+ <deck>.pptx compiled deck
140
+ ```
141
+
142
+ ## Compare strip
143
+
144
+ Build one when the user asks to see options, or the brief leaves the look open on purpose ("어떤
145
+ 느낌이 좋을지 모르겠어요", "show me a couple of styles"). Take the deck's frontmatter and its first
146
+ three slides, compile them under the chosen pack and both alternatives, render, and show three
147
+ labelled rows before writing the rest. Content and order are identical across rows; only
148
+ `tonality:` changes.
149
+
150
+ ```bash
151
+ for t in night ledger paper; do
152
+ sed "s/^tonality: .*/tonality: $t/" strip.md > "strip-$t.md"
153
+ node "$OFFICE" pptx compile "strip-$t.md" --pptx "strip-$t.pptx"
154
+ soffice -env:UserInstallation="file://$PWD/.lo" --headless --convert-to pdf "strip-$t.pptx"
155
+ done
156
+ ```
157
+
158
+ Do not lobby for a row in the message beyond the card's reason. When the user picks, rewrite the
159
+ card's reason as "chosen by the user from the compare strip" and build the deck under that pack.
160
+
161
+ ## Named legacy templates
162
+
163
+ AZURE-PRO, AZURE-A2Z, BOILERPLATE-PRETENDARD and BOILERPLATE-A2Z keep their single look. They are
164
+ used only when the user names one or an existing source already carries `template:`; they are
165
+ never picked by the table. The card still gets written (Tonality: the template name), the reply
166
+ still offers two tonalities, and the gate downgrades the variety and median-band checks to
167
+ advisories, since asking for a template by name means asking for its one look.
168
+
169
+ ## What goes wrong
170
+
171
+ - Picking by topic ("a startup, so Signal") without counting the outline.
172
+ - A reply that names one look and no alternatives, or names none.
173
+ - A reason that repeats the deck type instead of naming the signals.
174
+ - A dial moved by three or more without the user's word.
175
+ - Forcing a slide into a family the pack lacks; the compile notes the substitution, and the fix is
176
+ another family for that slide or another pack.
177
+ - Swapping the tonality after all slides are built without updating the card and the reply.
178
+ - A card whose "wrong if" line could never be checked ("if the user dislikes it").
179
+
180
+ ## Example library
181
+
182
+ Every file in `examples/` compiles under its tonality and passes `node "$OFFICE" pptx qa`. Read the
183
+ closest one before writing a new deck. Data in all of them is invented and marked as such.
184
+
185
+ | File | Deck type | Tonality | Alternatives |
186
+ |---|---|---|---|
187
+ | `01-pitch-signal-ko.md` | pitch | Signal | Studio, Atlas |
188
+ | `02-pitch-studio-en.md` | pitch for a physical product | Studio | Signal, Atlas |
189
+ | `03-business-review-ledger-ko.md` | business review | Ledger | Gazette, Night |
190
+ | `04-research-talk-paper-en.md` | research talk | Paper | Night, Chalk |
191
+ | `05-research-keynote-night-ko.md` | research talk in a keynote hall | Night | Paper, Chalk |
192
+ | `06-lecture-chalk-ko.md` | lecture | Chalk | Paper, Studio |
193
+ | `07-data-review-night-en.md` | data-heavy review | Night | Ledger, Paper |
194
+ | `08-field-report-atlas-ko.md` | image-led field report | Atlas | Studio, Signal |
195
+ | `09-briefing-gazette-ko.md` | Korean text briefing | Gazette | Ledger |
196
+ | `10-status-update-ledger-en.md` | status update | Ledger | Gazette, Night |
197
+ | `11-portfolio-studio-ko.md` | image-led portfolio | Studio | Atlas, Signal |
198
+ | `12-tutorial-chalk-en.md` | technical tutorial | Chalk | Paper, Night |
199
+ | `13-status-update-gazette-ko.md` | status update read at a desk | Gazette | Ledger, Night |
200
+
201
+ Examples that place pictures reference `assets/example-a.png`, `assets/example-b.png` and
202
+ `assets/example-c.png` next to the source. Put real images under those names, or change the paths,
203
+ before compiling one of them.
@@ -0,0 +1,185 @@
1
+ ---
2
+ tonality: signal
3
+ title: 동네 약국 재고 공유 플랫폼 투자 제안
4
+ date: 2026-10-05
5
+ department: 예시 헬스케어 스타트업
6
+ presenter: 예시 대표
7
+ notice: 예시 데이터 — 실제 수치로 바꿔 주세요
8
+ ---
9
+
10
+ ---
11
+ layout: cover-split-field
12
+
13
+ # 동네 약국 재고 공유 플랫폼 시드 투자 제안
14
+ ---
15
+
16
+ ---
17
+ layout: big-number
18
+
19
+ ## 동네 약국 유효기간 경과 의약품 폐기 규모
20
+
21
+ | 연간 폐기액 (약국당) | 폐기 품목 중 전문약 | 반품 불가 비중 | 약국 수 (수도권) |
22
+ |---|---|---|---|
23
+ | 310만 원 | 72% | 41% | 1만 1천 곳 |
24
+ | 2025년 표본 평균 | 금액 기준 | 낱알·개봉 포장 | 2025년 말 기준 |
25
+
26
+ 출처: 예시 약사회 재고 실태 조사, 수도권 약국 220곳, 2025년 11월 (예시)
27
+
28
+ > 수도권 약국 재고 조사, 2025년 (예시)
29
+
30
+ - 약국 한 곳이 해마다 약 310만 원어치 약을 기한이 지나 버린다
31
+ - 폐기액의 72%는 처방이 드문 전문약이고, 그중 41%는 도매상 반품이 안 된다
32
+ - 같은 약이 2km 안의 다른 약국에서는 품절인 경우가 표본의 38%였다
33
+ - 주: 폐기액은 약가 기준, 마약류와 냉장 보관 품목은 제외
34
+ ---
35
+
36
+ ---
37
+ layout: quote
38
+
39
+ ## 시범 참여 약사 인터뷰
40
+
41
+ “처방전 한 장 때문에 한 통을 들였다가 남은 스물여덟 알을 버리는 일이 한 달에 몇 번씩 있었습니다.”
42
+
43
+ — 예시 약국 대표 약사, 시범 운영 3개월 차 인터뷰 (2026년 7월)
44
+ ---
45
+
46
+ ---
47
+ layout: image-split
48
+
49
+ ## 근처 약국 재고 조회와 거래 화면
50
+
51
+ ![도 1. 재고 조회 화면 예시 | 출처: 예시 이미지](assets/example-a.png)
52
+
53
+ 출처: 시범 서비스 운영 기준, 2026년 9월 (예시)
54
+
55
+ - **조회** 반경 3km 안 참여 약국의 남은 수량과 유효기간을 품목별로 본다
56
+ - **요청** 필요한 낱알 수를 적어 보내면 상대 약국이 30분 안에 수락한다
57
+ - **정산** 약가의 90%로 거래하고 플랫폼 수수료 5%를 뗀다
58
+ - **배송** 오후 2시 전 요청은 당일 오토바이 퀵으로 받는다
59
+ - **기록** 거래 내역이 의약품 관리 대장 양식으로 자동 저장된다
60
+ ---
61
+
62
+ ---
63
+ layout: process
64
+
65
+ ## 재고 등록부터 정산까지 거래 단계
66
+
67
+ - **1. 등록** 약국 관리 프로그램의 재고 파일을 하루 한 번 올리면 기한 90일 이내 품목이 자동으로 공개된다
68
+ - **2. 요청** 품절 약국이 수량과 수령 시각을 정해 요청하고, 공개 약국 세 곳에 동시에 알림이 간다
69
+ - **3. 인계** 먼저 수락한 약국이 포장해 두면 퀵 기사가 수거하고, 받는 약국이 바코드로 수량을 확인한다
70
+ - **4. 정산** 매달 말 약국별 판매·구매 금액을 상계해 차액만 다음 달 10일에 이체한다
71
+ - 주: 시범 운영 거래 기록 2,640건 기준(2026년 4-9월, 예시), 등록 파일은 약국 관리 프로그램 3종의 표준 형식만 받는다
72
+ ---
73
+
74
+ ---
75
+ layout: kpi-row
76
+
77
+ ## 시범 운영 6개월 거래 지표
78
+
79
+ | 참여 약국 | 월 거래 | 요청 수락률 | 평균 인계 시간 | 약국당 폐기 감소 |
80
+ |---|---|---|---|---|
81
+ | 186곳 | 610건 | 83% | 2시간 40분 | 월 9만 원 |
82
+ | 4월 42곳 | 4월 120건 | 목표 80% | 퀵 배송 기준 | 참여 전 3개월 대비 |
83
+
84
+ 출처: 시범 서비스 거래 기록과 참여 약국 폐기 대장, 2026년 4-9월 (예시)
85
+
86
+ > 2026년 9월 기준, 시범 지역 3개 구 (예시)
87
+
88
+ - 참여 약국은 6개월 동안 42곳에서 186곳으로 늘었다
89
+ - 요청 열 건 중 여덟 건 이상이 수락되어 목표를 넘었다
90
+ - 폐기 감소액은 월 9만 원으로, 연간 폐기액의 약 3분의 1에 해당한다
91
+ - 주: 폐기 감소는 참여 전 3개월 평균과 9월 실적의 차이
92
+ ---
93
+
94
+ ---
95
+ layout: chart-insight
96
+
97
+ ## 월별 거래 건수, 2026년 4-9월
98
+
99
+ ::: chart type=column unit="건"
100
+ | 월 | 거래 건수 (건) |
101
+ |---|---|
102
+ | 4월 | 120 |
103
+ | 5월 | 210 |
104
+ | 6월 | 330 |
105
+ | 7월 | 420 |
106
+ | 8월 | 520 |
107
+ | 9월 | 610 |
108
+ > 월별 거래 건수, 시범 지역 3개 구, 시범 서비스 거래 기록 (예시 데이터)
109
+ :::
110
+
111
+ - 거래는 매달 늘었고 증가 폭은 7월 이후 월 90-100건으로 일정해졌다
112
+ - 약국 한 곳의 월 거래는 4월 2.9건에서 9월 3.3건으로 소폭 늘었다
113
+ - 광고 없이 약사회 지부 설명회 두 번으로 참여 약국을 모았다
114
+ - 주: 거래 건수는 인계까지 끝난 건만 셈, 취소 건 제외
115
+ ---
116
+
117
+ ---
118
+ layout: comparison
119
+
120
+ ## 도매상 반품과 약국 간 공유 비교
121
+
122
+ :::: columns 1fr 1fr
123
+ ::: col
124
+ - **도매상 반품**
125
+ (1) 회수액: 개봉하지 않은 통만 받고, 약가의 60-70%를 돌려준다
126
+ (2) 소요 기간: 반품 신청부터 정산까지 평균 5주가 걸린다
127
+ (3) 대상 품목: 낱알과 개봉 포장은 받지 않아 폐기액의 41%가 빠진다
128
+ (4) 품절 대응: 다음 정기 배송까지 하루에서 사흘을 기다린다
129
+ :::
130
+ ::: col
131
+ - **약국 간 공유**
132
+ (1) 회수액: 낱알 단위로 팔고 약가의 85.5%를 받는다 (수수료 차감 후)
133
+ (2) 소요 기간: 수락 후 평균 2시간 40분에 인계, 정산은 다음 달 10일
134
+ (3) 대상 품목: 기한 90일 이내 품목과 개봉 포장까지 거래할 수 있다
135
+ (4) 품절 대응: 같은 날 오후에 근처 약국에서 받아 조제한다
136
+ :::
137
+ ::::
138
+
139
+ - 주: 예시 도매상 반품 약관과 시범 서비스 거래 기록(2026년) 기준
140
+ ---
141
+
142
+ ---
143
+ layout: table-insight
144
+
145
+ ## 거래 1건당 수익 구조, 2026년 9월
146
+
147
+ | 항목 | 금액 (원) | 비중 | 4월 (원) |
148
+ |---|---|---|---|
149
+ | 평균 거래액 | 48,000 | 100% | 41,000 |
150
+ | 판매 약국 정산 | 43,200 | 90% | 36,900 |
151
+ | 플랫폼 수수료 | 2,400 | 5% | 2,050 |
152
+ | 퀵 배송비 (받는 약국 부담) | 4,500 | — | 5,200 |
153
+ | 결제·서버 비용 | 900 | 2% | 1,100 |
154
+ | 건당 공헌이익 | 1,500 | 3% | 950 |
155
+
156
+ 출처: 시범 서비스 정산 자료, 2026년 4월·9월 (예시)
157
+
158
+ > 표 1. 거래 1건당 수익 구조, 2026년 9월 평균 (예시 데이터)
159
+
160
+ - 수수료 2,400원에서 결제·서버 비용을 빼면 건당 1,500원이 남는다
161
+ - 월 거래가 3,000건을 넘으면 운영 인력 비용을 공헌이익으로 메울 수 있다
162
+ - 배송비는 묶음 배송 시범으로 건당 4,500원에서 3,200원까지 낮추는 중이다
163
+ - 주: 공헌이익은 수수료에서 결제·서버 비용을 뺀 금액, 인건비 제외
164
+ ---
165
+
166
+ ---
167
+ layout: closing-ask
168
+
169
+ ## 시드 투자 8억 원 사용 계획
170
+
171
+ | 쓰임 | 금액 | 시점 | 담당 | 목표 |
172
+ |---|---|---|---|---|
173
+ | 수도권 12개 구로 지역 확대 | 3억 2천만 원 | 2027년 상반기 | 사업팀 | 참여 약국 900곳 |
174
+ | 약국 관리 프로그램 연동 3종 추가 | 2억 원 | 2026년 4분기 | 개발팀 | 등록 시간 하루 1분 이하 |
175
+ | 묶음 배송 운영 | 1억 2천만 원 | 2027년 1분기 | 운영팀 | 건당 배송비 3,200원 |
176
+ | 약사회 지부 제휴 | 8천만 원 | 상시 | 제휴팀 | 지부 설명회 24회 |
177
+ | 운영 예비비 | 8천만 원 | 필요 시 | 재무팀 | 4개월 비용 완충 |
178
+ | 합계 | 8억 원 | 18개월 | — | — |
179
+
180
+ > 표 2. 투자금 쓰임 (예시 데이터)
181
+
182
+ - 요청: 18개월 운영을 위한 시드 투자 8억 원
183
+ - 기대 성과: 18개월 뒤 월 거래 4,000건, 월 공헌이익 600만 원
184
+ - 다음 단계: 10월 넷째 주 실사 자료 공유, 담당 예시 재무 책임자
185
+ ---