academic-slides
Create Beamer-inspired academic HTML presentations from scratch or by converting PowerPoint files.
Skill by Nero1688
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/nero1688-academic-slides
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: presentations, frontend, research
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md and code. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/academic-slides
What's inside
- Skill name
academic-slides- Description
- Create Beamer-inspired academic HTML presentations from scratch or by converting PowerPoint files. Use when the user wants to build slides for a conference talk, lecture, seminar, or thesis defense. Supports theorem environments, KaTeX equations, algorithm pseudocode, and citations. Helps academics discover their preferred visual theme through Beamer-style previews rather than abstract choices.
- License
- MIT
- Instructions
- 511 lines in SKILL.md · view on GitHub
Agent Skills spec issues
- body: 511 lines (spec recommends under 500)
- files: referenced but missing: references/academic-style-catalog.md
Files (9)
- LICENSE
- README.md
- SKILL.md
- STYLE_PRESETS.md
- css-architecture.md · references/
- html-architecture.md · references/
- ppt-conversion.md · references/
- style-and-animation.md · references/
- troubleshooting-and-examples.md · references/
SKILL.md (first lines)
# Academic Slides Skill
Create zero-dependency, Beamer-inspired HTML presentations that run entirely in the browser. This skill helps academics and researchers build professional presentations with proper theorem environments, equation rendering via KaTeX, and structured frame navigation. Users discover their preferred theme through visual exploration ("show, don't tell"), then the skill generates production-quality frame decks.
## Core Philosophy
1. **Zero Build Dependencies** -- Single HTML files with inline CSS/JS. KaTeX loaded via CDN for equations. No npm, no build tools.
2. **Show, Don't Tell** -- People don't know what they want until they see it. Generate visual theme previews for academic users.
3. **Academic Authenticity** -- Beamer-inspired structure, Computer Modern typography, proper theorem environments. Not PowerPoint-with-serif-fonts.
4. **Production Quality** -- Accessible, responsive, well-commented code.
5. **Viewport Fitting (CRITICAL)** -- Every frame MUST fit exactly within the viewport. No scrolling within frames, ever. This is non-negotiable.
---
## CRITICAL: Viewport Fitting Requirements
**This section is mandatory for ALL presentations. Every frame must be fully visible without scrolling on any screen size.**
### The Golden Rule
```text
Each frame = exactly one viewport height (100vh/100dvh)
Content overflows? -> Split into multiple frames or reduce content
Never scroll within a frame.
```
### Content Density Limits
To guarantee viewport fitting, enforce these limits per frame:
| Frame Type | Maximum Content |
|------------|-----------------|
| Title Frame | 1 title + 1 subtitle + author/institute/date block |
| Content Frame | 1 heading + 4-5 bullet points OR 1 heading + 2 short paragraphs |
| Theorem/Proof Frame | 1 theorem box (max 4 lines) + 1 proof sketch (max 5 lines) |
| Equation Frame | 1 heading + 1-3 display equations with optional annotation |
| Algorithm Frame | 1 heading + max 12 lines of pseudocode |
| Definition Frame | 1 heading + 1-2 definition boxes (max 3 lines each) |
| Citation/References Frame | 1 heading + max 8 reference entries |
| Section Divider Frame | 1 section number + 1 section title + optional outline |
**If content exceeds these limits -> Split into multiple frames**
### CSS 架構與溢版防治(已拆出)
撰寫任何 slide HTML/CSS 前,讀 `references/css-architecture.md`:必用的 CSS 骨架、溢版防治檢查清單、內容塞不下時的處理順序、視口貼合測試法。上方的 Golden Rule 與密度上限是鐵律,細節與程式碼全在該檔。
## Phase 0: Detect Mode
First, determine what the user wants:
**Mode A: New Academic Presentation**
- User wants to create slides from scratch for a talk, lecture, or defense
- Proceed to Phase 0.5 (Essential Content Questions), then Phase 1 (Content Discovery)
**Mode B: PPT/PDF/Paper Conversion**
- User has an existing PowerPoint, PDF, or paper URL to convert to web slides
- Proceed to Phase 4 (Content Extraction), then Phase 0.5, then Phase 1P (Paper Focus Discovery)
**Mode C: Existing Presentation Enhancement**
- User has an HTML presentation and wants to improve it
- Read the existing file, understand the structure, then enhance
### Fast-Path Detection
If the user's initial message specifies enough to skip format questions (e.g., topic, approximate length, and/or theme name are all present), you MAY skip the format-mechanical questions in Phase 1 (purpose, length, content types, field). However, you MUST still run Phase 0.5 (Essential Content Questions) before generating. **No presentation should ever be generated without asking the Phase 0.5 questions.**
**Examples of fast-path triggers:**
- "/academic-slides 10-slide lecture on dynamic programming using Metropolis"
- "/academic-slides conference talk on our ICML paper, Berlin theme"
- "Make 15 slides about transformer attention for my lab meeting"
In all of these cases, format is clear from context but content quality questions have not been answered. Proceed to Phase 0.5.
---
## Phase 0.5: Essential Content Questions (MANDATORY)
**These questions MUST be asked before generating any presentation, regardless of mode or how much information the user provided in their initial message.** They cannot be inferred from topic or theme -- they require the user's judgment. Ask all 3 in a single AskUserQuestion call.
### Why These Questions Matter
A presentation on "transformer attention" for a room of NLP researchers looks completely different from one for undergraduate CS students. A talk where the takeaway is "our method is 3x faster" structures differently from one where the takeaway is "this theoretical framework unifies prior work." These questions determine content strategy, not just format.
### The Questions (Single AskUserQuestion Call)
**Question 1: Audience**
- Header: "Audience"
- Question: "Who is your audience?"
- Options:
- "Domain experts" -- They know the field well; skip basics, focus on contributions and technical details
- "Technical but not specialist" -- They have technical background but are not in your exact subfield
- "Students / newcomers" -- They need context, definitions, and motivation before results
- "Mixed / general academic" -- Varies widely; balance accessibility with depth
**Question 2: Key Takeaway**
- Header: "Takeaway"
- Question: "If the audience remembers ONE thing from your talk, what should it be?"
- Options:
- "A specific result" -- A theorem, benchmark, or experimental finding (will ask you to state it)
- "A new method or technique" -- How to do something they could not do before
- "A conceptual framework" -- A way of thinking about a problem or connecting ideas
- "A problem or open question" -- Motivating future work or collaboration
**Question 3: Narrative Arc**
- Header: "Structure"
- Question: "How should the presentation flow?"
- Options:
- "Problem -> Solution -> Results" -- Classic research talk structure
- "Background -> Our Work -> Impact" -- Contextualizes contribution in the field
- "Survey / Comparison" -- Review multiple approaches, compare trade-offs
- "Build-up / Tutorial" -- Teach concepts progressively, layer by layer
### Follow-Up: Key Result Statement
If the user chose "A specific result" as their takeaway, follow up immediately:
"Please state the key result in one sentence. This will be featured prominently in the slides."From Nero1688/claude-academic-skills, MIT.
More from Nero1688/claude-academic-skills
- 0.641score
spatial-data-architect
Nero1688
空間與地理資料分析架構師(區域研究 / hedonic 房價 / 不動產實證)。把帶有地址或經緯度的資料變成可分析、可畫圖、可發表的空間研究資料集。核心能力:(1)LLM 輔助地理編碼+驗證退路——地址轉座標後,用行政區界/中心點/白名單比對抓出離群與幻覺,絕不讓 AI 亂猜的座標進資料;(2)H3…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
bilingual-paper-reader
Nero1688
英文論文的雙欄對照精讀工具。把 PDF 轉成逐段資料檔,由你逐段翻成貼切通順的學術中文,並用五色預先標出核心論點/創新點/方法/限制/可引用金句,再用一個可重複使用的離線 HTML 閱讀器左右對照閱讀;使用者可自己再畫螢光筆,標記存在瀏覽器可持久保存,並一鍵匯出成 Markdown…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
academic-deck-animator
Nero1688
學術簡報動畫引擎:以 slides_content.json 統一規格驅動的雙引擎動態簡報產生器。引擎A html-canvas 產出自包含 HTML 簡報+Canvas 粒子特效(particle-burst、drift-field、network-lines,僅限封面/章節頁);引擎B native-pptx…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
global-opendata-scout
Nero1688
跨國公開統計資料偵察員。研究要跨國比較、或使用者不在台灣時,判斷該用哪個國際資料庫、哪個國家的統計機構、能不能程式化撈取。內建已實測可用的免金鑰端點(World Bank、Eurostat、ILOSTAT、IMF、UN…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
journal-submission-scout
Nero1688
投稿期刊選擇顧問。稿子寫好了要投哪裡?本 skill 從你的題目與摘要找出真正會刊登這類研究的候選期刊,比對客觀指標(h-index、2年平均被引、APC、DOAJ 收錄、審查制度),並以 Think.Check.Submit 架構做**掠奪性期刊篩查**——這是本 skill…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
tej-variable-mapper
Nero1688
把國際期刊(Compustat / CRSP)的變數操作性定義,精準映射到 TEJ 台灣經濟新報的欄位與代碼。輸出三欄對照表:期刊定義(原公式+原代碼如 AT、Item 6)/TEJ 欄位與代碼/GAAP-IFRS 落差與代理方案。內建『假說→變數→資料欄位』對帳表;映射不確定的欄位標『推測+驗證方法』,絕不編造…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
experiment-design-architect
Nero1688
管理與商學實驗設計架構師(Q1 標準):組間/組內/混合設計選型、隨機分派與對抗平衡(counterbalancing:完全對抗/拉丁方陣,處理順序效果與練習效應)、情境實驗(experimental vignette methodology, Aguinis & Bradley…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
phd-milestone-tracker
Nero1688
商管博士班修業里程碑與時程追蹤器,專為在職生設計。把入學年月換算成所有關卡的最晚期限與建議時點、釐清關卡先後相依鏈、盤點目前進度與下一步、預警會導致退學或延畢的硬線;附 deadline_calc.py…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
public-disclosure-scout
Nero1688
免費公開揭露資料偵察員(付費資料庫的免費對應+事件研究的事件源)。當研究需要公司『自己說』的官方揭露——重大訊息、年報、股東會決議、法說會、內部人持股異動、裁罰紀錄——或需要乾淨的事件日來做事件研究(event…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
qual-exam-coach
Nero1688
商管博士班『學科資格考』備考教練,專為策略組織與人資組設計。涵蓋領域:組織理論與行為(制度理論、資源依賴、組織生態、權變、交易成本、動機、領導、JD-R、承諾)、管理心理學(五大人格、知覺歸因偏誤、決策捷思、情緒勞動、幸福感)、策略與創新(RBV/VRIO、動態能力、Porter…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
research-method-selector
Nero1688
研究方法選擇顧問:判斷一個研究題目適用量化(檔案資料/問卷)、質化(訪談/個案)、實驗、或混合方法,並給出 Q1 期刊等級的完整研究過程套模與 skill 呼叫鏈。核心邏輯:Edmondson & McManus (2007) 方法論適配(methodological…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
defense-qa-coach
Nero1688
學位口試與研討會問答的『答辯教練』。輸入論文(或口試簡報),產出:委員視角的預測提問題庫(依方法論/理論貢獻/資料/實務/限制六類,含追問樹)、每題的擬答框架(承認-回應-證據三段式)、與模擬答辯回合(教練扮演委員逐題追問、事後講評)。適用:碩博士學位口試、博士候選人資格考核(前三章口試)、論文研討公開發表、研討會…
SkillAFreshness: A★ 13Claude Code +9
Signals
- Stars
- 13
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Sep 20, 2026 (1 month ago)
- Last release
- Oct 22, 2026 (7 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from Nero1688
- 0.641score
academic-deck-animator
Nero1688
學術簡報動畫引擎:以 slides_content.json 統一規格驅動的雙引擎動態簡報產生器。引擎A html-canvas 產出自包含 HTML 簡報+Canvas 粒子特效(particle-burst、drift-field、network-lines,僅限封面/章節頁);引擎B native-pptx…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
phd-milestone-tracker
Nero1688
商管博士班修業里程碑與時程追蹤器,專為在職生設計。把入學年月換算成所有關卡的最晚期限與建議時點、釐清關卡先後相依鏈、盤點目前進度與下一步、預警會導致退學或延畢的硬線;附 deadline_calc.py…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
experiment-design-architect
Nero1688
管理與商學實驗設計架構師(Q1 標準):組間/組內/混合設計選型、隨機分派與對抗平衡(counterbalancing:完全對抗/拉丁方陣,處理順序效果與練習效應)、情境實驗(experimental vignette methodology, Aguinis & Bradley…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
global-opendata-scout
Nero1688
跨國公開統計資料偵察員。研究要跨國比較、或使用者不在台灣時,判斷該用哪個國際資料庫、哪個國家的統計機構、能不能程式化撈取。內建已實測可用的免金鑰端點(World Bank、Eurostat、ILOSTAT、IMF、UN…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
journal-submission-scout
Nero1688
投稿期刊選擇顧問。稿子寫好了要投哪裡?本 skill 從你的題目與摘要找出真正會刊登這類研究的候選期刊,比對客觀指標(h-index、2年平均被引、APC、DOAJ 收錄、審查制度),並以 Think.Check.Submit 架構做**掠奪性期刊篩查**——這是本 skill…
SkillAFreshness: A★ 13Claude Code +9 - 0.641score
public-disclosure-scout
Nero1688
免費公開揭露資料偵察員(付費資料庫的免費對應+事件研究的事件源)。當研究需要公司『自己說』的官方揭露——重大訊息、年報、股東會決議、法說會、內部人持股異動、裁罰紀錄——或需要乾淨的事件日來做事件研究(event…
SkillAFreshness: A★ 13Claude Code +9
Cite this page
academic-slides skill. Agent Mart. https://agentmart.dev/skills/nero1688-academic-slides (last updated 2026-10-24)