SuperAPI AI Directory

scholar-stickers

Office & productivity

View repository

Updated 10-22 19:36

About this skill

Scholar Stickers · Knowledge Character Sticker Slide Generator

Turn a group of figures (thinkers / scientists / literary figures / historical figures) or a group of concepts intoa single-file 16:9 interactive slide deck: overview page sticker wall → click a sticker → figure detail page (biography expandable on click). Intended for teacher classroom projection,zero scrolling, keyboard paging, readable from the back row。


1. Input parameters

Parameter Required Description Example
figures ✅ List of 3–6 figures Socrates, Confucius, Rousseau, Dewey, Piaget
domain ✅ Subject area (determines the direction of research) Philosophy of education / Classical physics / Tang and Song literature
title ✅ Overview page main title Philosophy Begins with Wonder and Doubt
subtitle ⬜ Overview page introduction (empty by default) Why do we educate…
chapter ⬜ Top chapter number (default CHAPTER · 01) CHAPTER · 03
theme ⬜ Color theme (default warm-classic, see §5.2) ink-blue / forest
expand_mode ⬜ Whether biography expands on click (on by default) on / off

If the user gives only a list of figures without title/domain,first draft 2–3 titles based on the field and let the user choose, do not make the final decision on your own.


2. Workflow (five steps)

Step 1 · Material organization (five sections per person)

Use WebSearch / a local knowledge base to verify and organize the following fields for each figure.All fields have hard character limits(if it does not fit on one screen, it fails):

{
  id: 'socrates',            // 拼音 id,用于图片文件名
  name: '苏格拉底',           // 中文姓名
  nameEn: 'Socrates',        // 外文名(中国人物用拼音或字)
  years: '470 BC – 399 BC',  // 生卒年
  keywords: '古希腊 · 街头追问 · 产婆术',   // ≤4 个关键词,· 分隔
  bio: '……约 110 字……',     // 生平:4-5 行以内,3-5 个关键事件
  contribs: [                // 贡献:恰好 3 条,每条 ≤40 字,<b>加粗术语</b>
    '<b>术语</b>:一句话解释。',
    '<b>术语</b>:一句话解释。',
    '<b>术语</b>:一句话解释。'
  ],
  quotes: ['……', '……'],     // 金句:恰好 2 条,每条 ≤30 字
  books: ['《A》', '《B》'],   // 经典文本:恰好 2 本
  inspiration: '……≤22 字,一行说完。'   // 当代启示:单句
}

Content red lines:

  • Mark uncertain historical facts with [待核实], do not fabricate birth/death years, work titles, or quotations
  • Famous quotes must besourced original text or widely accepted paraphrases, if none can be found, replace with a famous line from their works
  • For Chinese figures, quotes should preferably use original classical texts (classical Chinese may be retained; an optional plain-language gloss may be added)

Step 2 · Portrait generation (pencil sketch sticker style)

Use the current platform'stext-to-imagecapability (WorkBuddy uses the "multimodal content generation" skill; other platforms use their built-in image generation tools) to generate a portrait for each figure in a unified style.

Prompt template(replace {描述}):

Pencil sketch portrait of {人物身份描述:国籍/时代/外貌特征/服装},
graphite pencil drawing on white paper, monochromatic black and white,
hand-drawn white sticker border around the portrait with slightly torn paper edges,
bust composition, classical engraving style

The appearance description should includeidentifiable features(beard/glasses/hat/hairstyle/clothing), for example:

  • Socrates → ancient Greek philosopher Socrates, bald head with curly white side hair, long curly white beard, classical Greek toga draped over one shoulder, intense thoughtful gaze
  • Dewey → American philosopher John Dewey, round wire-rimmed glasses, small mustache, neat combed gray hair, formal suit with tie, dignified expression

Calling tips(platform-general + WorkBuddy tested):

  • General principles: just use the current platform/tool's built-in text-to-image capability, use the template above for all prompts, output 768×960 (or a close vertical ratio), and save PNGs to the portraits/ subdirectory of the delivery directory
  • WorkBuddy (tested 2026-08):
    1. First call connect_cloud_service Get tempToken (re-fetch for each task)
    2. Script: locate the resources/app.asar.unpacked/resources/builtin-skills/buddy-multimodal-generation/scripts/buddy-cloud.pyunder the local WorkBuddy installation directory (use where buddy-cloud.py or search by installation path; do not hardcode the username)
    3. Python: prefer python3; if not found, use py -3 or where python to probe (do not hardcode absolute paths)
    4. Parameters:image "<prompt>" --resolution 768:960 --token-stdin
    5. Pass the Token via a pipe:echo -n "<token>" | <python> -u <script> image "..." --token-stdin
    6. After generation, curl -sS -L -o portraits/{id}.png "<result_url>" lands in the portraits/ subdirectory of the delivery directory
    7. If it reports "250 task limit", wait 30 seconds and retry once; retry at most 1 time
  • After generationyou must Read and check each one: whether the person is recognizable, and whether there is a watermark or corner badge (if watermarked, crop it out or regenerate)

Step 3 · Page assembly

Copy template.html(in this skill directory) to the delivery directory and replace:

  1. THEME object → color scheme (see §5.2)
  2. META object → title / subtitle / chapter
  3. FIGURES array → Step 1 data
  4. Confirm that portraits/ relative-path images are all present

Step 4 · Self-check list

  • On the overview page, the sticker wall is arranged 3×2 without misalignment, and hovers back to level and floats up
  • Each detail page has all 4 rows: sticker row / biography (collapsed by default) / contributions / text + insights
  • The biography click-to-expand animation works normally, and it automatically resets to collapsed after page turns
  • Keyboard ←→ turns pages, Esc returns to overview, bottom dots are clickable
  • When the window is resized, the stage always stays 16:9 centered with no scrolling
  • Quotes 28px, body text 22px—readable when projected
  • No content overflow (bio ≤5 lines, each contribs item within one or two lines)

Step 5 · Delivery

preview_url Open the HTML preview → deliver_attachments Deliver the HTML + all portrait PNGs + content document MD (optional).


3. Page structure (fixed, must not be changed arbitrarily)

SLIDE 0  总览页
├── 顶部条:CHAPTER · NN | BEAT NN(期刊式编号)
├── 左:eyebrow + 大标题(em 斜体朱红强调词)+ 引言 + 点击提示
└── 右:贴纸墙(3 列 × 2 行,各自轻微旋转,hover 回正)

SLIDE 1..N  人物页(每人一页,纵向 4 行)
├── 第 1 行「贴纸行」:肖像(240px 白边贴纸) + 名字块(中64/EN26/年代21/关键词16.5) + 金句块(28px 大引号)
├── 第 2 行「生平」:标题条(点击展开 ▸) + 正文(22px,默认收起)
├── 第 3 行「贡献」:标题 + 3 条 ◆ 列表(22px)
└── 第 4 行「文本+启示」:书签(18px × 2) | ✦ 启示条(20px 单行)

底部条(每页):导航圆点 | 键盘提示 | 页码 NN / MM

Interaction: click a sticker to jump to details / ←→ · Space · PgUpPgDn to turn pages / Esc · Home to return to overview / End to go to the last page / click the biography to expand (automatically resets on page turn).


4. Design specifications (classroom readability first)

4.1 Layout

  • 16:9 fixed stage: design canvas 1600×900, JS fitStage() calculates scale = min(vw/1600, vh/900),transform: scale() centered; outer layer has a dark background. On any screen it is a standard full-screen 16:9 with zero scrolling.
  • Vertical single-column flow detail page:Do not use "left = person, right = matter"; use a top-to-bottom flow—the teacher's gaze follows a single main axis while lecturing.

4.2 Font size baseline (1600×900 design canvas px, readable baseline for the back row)

Level Font size Font weight
Overview main title 80 900
Person name (Chinese) 64 900
Quote 28 600
Biography / contributions body 22 400
Insights 20 400
section title 19 700
Bookmark 18 400
English name / era 26 / 21 —
Keywords 16.5 400

When increasing font size, tighten line-height slightly in reverse (22px with 1.7) to maintain density;section titles must scale up in sync with the body text, as they are the visual anchors.

4.3 Fonts

'Cormorant Garamond'(拉丁/斜体/数字) + 'Noto Serif SC'(中文标题/正文) + 'Noto Sans SC'(中文辅助)

Google Fonts CDN import; the bilingual Chinese-English labels are a fixed style (BIOGRAPHY · 生平事迹).

4.4 Color Themes

Default warm-classic(visual signature of this skill):

Token Value Usage
--bg #f3ebdb Page warm beige background
--paper #fbf6e8 Quote block / paper surface
--ink #1a1612 Main text
--vermillion #b8392a Vermilion accent (title words / quotation marks / arrows / dots)
--ink-mute #6a5e4a Secondary text

Alternative themes (just change the :root variable):

  • ink-blue: vermillion→#2c5a7a, bg→#e9edf2 (cool academic)
  • forest: vermillion→#3a6b4f, bg→#eaf0e6 (natural sciences)
  • When the user specifies a brand color, replace only the two tokens --vermillion and --bg ; leave the rest unchanged.

4.5 Sticker Style

White background, padding 7px + three-layer shadow (near/mid/far) + staggered rotation from -2.5° to 2.5°; hover: rotate(0) translateY(-8px) scale(1.05).


5. Templates and Assets

  • template.html — parameterized full page (the three replacement points THEME / META / FIGURES are marked with comments at the top of the file)
  • Store all portraits in portraits/{id}.png(768×960)

6. Common Variants

Requirement Approach
Add a person Add an entry to FIGURES + generate a portrait; the overview wall automatically arranges in 3×N rows
Concept/principle page (non-person) Replace the portrait with a "principle diagram sticker" (generate a simple sketch diagram), and put a one-sentence statement of the principle in the quote slot
Development history timeline Add a timeline ribbon at the bottom of the overview page; or expand the period background in the years row on the detail page
No click-to-expand expand_mode: off: delete the .d-bio-content collapse CSS so p displays directly
All-English version Swap name/n ↔ nameEn, and reduce each font size by 2px

7. Red Lines

  • No fabrication: birth/death years, works, and quotes must be verifiable; if not verifiable, change the content, do not make it up
  • Do not exceed word limits: bio ≤120 characters, inspiration ≤22 characters, quotes ≤30 characters each — if over, cut content, not shrink the font size
  • No scrolling: under no circumstances may a scrollbar appear on the page (a capital offense for classroom projection)
  • AI portraits must have watermarks removed: crop them out or regenerate, and check each one before delivery