paper-slides

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Paper Slides: From Paper to Conference Talk

论文转幻灯片：从论文到会议报告

Generate conference presentation slides from: $ARGUMENTS

从以下内容生成会议演示幻灯片：$ARGUMENTS

Context

背景

This skill runs after Workflow 3 (

/paper-writing

). It takes a compiled paper and generates a presentation slide deck for conference oral talks, spotlight presentations, or poster lightning talks.

Unlike posters (single page, visual-first), slides tell a temporal story: each slide builds on the previous one, with progressive revelation of the research narrative. A good talk makes the audience understand why this matters before showing what was done.

本技能需在工作流3（

/paper-writing

）完成后运行。它会接收已编译的论文，为会议口头报告、焦点报告或海报闪电报告生成演示幻灯片套件。

与海报（单页、视觉优先）不同，幻灯片讲述的是时序化故事：每张幻灯片都基于前一张内容展开，逐步呈现研究叙事。一份优秀的报告应先让听众理解「这项研究为何重要」，再展示「我们做了什么」。

Constants

常量配置

VENUE =
NeurIPS
— Target venue, determines color scheme. Supported:
```
NeurIPS
```
,
```
ICML
```
,
```
ICLR
```
,
```
AAAI
```
,
```
ACL
```
,
```
EMNLP
```
,
```
CVPR
```
,
```
ECCV
```
,
```
GENERIC
```
. Override via argument.
TALK_TYPE =
spotlight
— Talk format. Options:
```
oral
```
(15-20 min),
```
spotlight
```
(5-8 min),
```
poster-talk
```
(3-5 min),
```
invited
```
(30-45 min). Determines slide count and content depth.
TALK_MINUTES = 15 — Talk duration in minutes. Auto-adjusts slide count (~1 slide/minute for oral, ~1.5 slides/minute for spotlight). Override explicitly if needed.
ASPECT_RATIO =
16:9
— Slide aspect ratio. Options:
```
16:9
```
(default, modern projectors),
```
4:3
```
(legacy).
SPEAKER_NOTES = true — Generate
```
\note{}
```
blocks in beamer and corresponding PPTX notes. Set
```
false
```
for clean slides without notes.
PAPER_DIR =
paper/
— Directory containing the compiled paper.
OUTPUT_DIR =
slides/
— Output directory for all slide files.
REVIEWER_MODEL =
gpt-5.4
— Model used via Codex MCP for slide review.
AUTO_PROCEED = false — At each checkpoint, always wait for explicit user confirmation.
COMPILER =
latexmk
— LaTeX build tool.
ENGINE =
pdflatex
— LaTeX engine. Use
```
xelatex
```
for CJK text.

💡 Override:

/paper-slides "paper/" — talk_type: oral, venue: ICML, minutes: 20, aspect: 4:3

VENUE =
NeurIPS
— 目标会议，决定配色方案。支持的会议：
```
NeurIPS
```
、
```
ICML
```
、
```
ICLR
```
、
```
AAAI
```
、
```
ACL
```
、
```
EMNLP
```
、
```
CVPR
```
、
```
ECCV
```
、
```
GENERIC
```
。可通过参数覆盖默认值。
TALK_TYPE =
spotlight
— 报告格式。可选值：
```
oral
```
（15-20分钟）、
```
spotlight
```
（5-8分钟）、
```
poster-talk
```
（3-5分钟）、
```
invited
```
（30-45分钟）。决定幻灯片数量和内容深度。
TALK_MINUTES = 15 — 报告时长（分钟）。会自动调整幻灯片数量（口头报告约1分钟/张，焦点报告约1.5分钟/张）。如有需要可手动覆盖。
ASPECT_RATIO =
16:9
— 幻灯片宽高比。可选值：
```
16:9
```
（默认，适配现代投影仪）、
```
4:3
```
（ legacy设备）。
SPEAKER_NOTES = true — 在beamer中生成
```
\note{}
```
块，并同步生成对应的PPTX备注。若需无备注的简洁幻灯片，可设置为
```
false
```
。
PAPER_DIR =
paper/
— 存放已编译论文的目录。
OUTPUT_DIR =
slides/
— 所有幻灯片文件的输出目录。
REVIEWER_MODEL =
gpt-5.4
— 通过Codex MCP用于幻灯片审核的模型。
AUTO_PROCEED = false — 在每个检查点，始终等待用户明确确认后再继续。
COMPILER =
latexmk
— LaTeX构建工具。
ENGINE =
pdflatex
— LaTeX引擎。若论文包含中日韩文本，使用
```
xelatex
```
。

💡 覆盖默认值示例：

/paper-slides "paper/" — talk_type: oral, venue: ICML, minutes: 20, aspect: 4:3

Talk Type → Slide Count

报告类型 → 幻灯片数量

Talk Type	Duration	Slides	Content Depth
`poster-talk`	3-5 min	5-8	Problem + 1 method slide + 1 result + conclusion
`spotlight`	5-8 min	8-12	Problem + 2 method + 2 results + conclusion
`oral`	15-20 min	15-22	Full story with motivation, method detail, experiments, analysis
`invited`	30-45 min	25-40	Comprehensive: background, related work, deep method, extensive results, discussion

报告类型	时长	幻灯片数量	内容深度
`poster-talk`	3-5分钟	5-8张	问题 + 1张方法页 + 1张结果页 + 结论
`spotlight`	5-8分钟	8-12张	问题 + 2张方法页 + 2张结果页 + 结论
`oral`	15-20分钟	15-22张	完整叙事：动机、方法细节、实验、分析
`invited`	30-45分钟	25-40张	全面覆盖：背景、相关工作、深度方法、大量结果、讨论

Venue Color Schemes

会议配色方案

Same as

/paper-poster

Venue	Primary	Accent	Background	Text
NeurIPS	`#8B5CF6`	`#2563EB`	`#FFFFFF`	`#1E1E1E`
ICML	`#DC2626`	`#1D4ED8`	`#FFFFFF`	`#1E1E1E`
ICLR	`#059669`	`#0284C7`	`#FFFFFF`	`#1E1E1E`
CVPR	`#2563EB`	`#7C3AED`	`#FFFFFF`	`#1E1E1E`
GENERIC	`#334155`	`#2563EB`	`#FFFFFF`	`#1E1E1E`

与

/paper-poster

一致：

会议	主色调	强调色	背景色	文字色
NeurIPS	`#8B5CF6`	`#2563EB`	`#FFFFFF`	`#1E1E1E`
ICML	`#DC2626`	`#1D4ED8`	`#FFFFFF`	`#1E1E1E`
ICLR	`#059669`	`#0284C7`	`#FFFFFF`	`#1E1E1E`
CVPR	`#2563EB`	`#7C3AED`	`#FFFFFF`	`#1E1E1E`
GENERIC	`#334155`	`#2563EB`	`#FFFFFF`	`#1E1E1E`

State Persistence (Compact Recovery)

状态持久化（快速恢复）

Persist state to

slides/SLIDES_STATE.json

after each phase:

json

{
  "phase": 3,
  "venue": "NeurIPS",
  "talk_type": "spotlight",
  "slide_count": 10,
  "codex_thread_id": "019cfcf4-...",
  "status": "in_progress",
  "timestamp": "2026-03-18T15:00:00"
}

On startup: if

SLIDES_STATE.json

exists with

"status": "in_progress"

and within 24h → resume. Otherwise → fresh start.

在每个阶段完成后，将状态保存到

slides/SLIDES_STATE.json

：

json

{
  "phase": 3,
  "venue": "NeurIPS",
  "talk_type": "spotlight",
  "slide_count": 10,
  "codex_thread_id": "019cfcf4-...",
  "status": "in_progress",
  "timestamp": "2026-03-18T15:00:00"
}

启动时：若存在

SLIDES_STATE.json

且

"status": "in_progress"

，且时间在24小时内，则恢复之前的工作流。否则，从头开始。

Workflow

工作流

Phase 0: Input Validation & Setup

阶段0：输入验证与环境准备

Check prerequisites:
bash
```
which pdflatex && which latexmk
```

Verify paper exists:

bash

ls $PAPER_DIR/main.tex || ls $PAPER_DIR/main.pdf
ls $PAPER_DIR/sections/*.tex
ls $PAPER_DIR/figures/

Backup existing slides: if
```
slides/
```
exists, copy to
```
slides-backup-{timestamp}/
```
Create output directory:
```
mkdir -p slides/figures
```
Detect CJK: if paper contains Chinese/Japanese/Korean, set ENGINE to
```
xelatex
```
Determine slide count: from TALK_TYPE and TALK_MINUTES using the table above
Check for resume: read
```
slides/SLIDES_STATE.json
```
if it exists

State: Write

SLIDES_STATE.json

with

phase: 0

检查前置依赖：
bash
```
which pdflatex && which latexmk
```

验证论文文件存在：

bash

ls $PAPER_DIR/main.tex || ls $PAPER_DIR/main.pdf
ls $PAPER_DIR/sections/*.tex
ls $PAPER_DIR/figures/

备份现有幻灯片：若
```
slides/
```
目录已存在，将其复制到
```
slides-backup-{timestamp}/
```
创建输出目录：
```
mkdir -p slides/figures
```
检测中日韩文本：若论文包含中文/日文/韩文，将ENGINE设置为
```
xelatex
```
确定幻灯片数量：根据TALK_TYPE和TALK_MINUTES，参考上方表格计算
检查是否恢复工作流：若存在
```
slides/SLIDES_STATE.json
```
则读取状态

状态保存：写入

SLIDES_STATE.json

，标记

phase: 0

。

Phase 1: Content Extraction & Slide Outline

阶段1：内容提取与幻灯片大纲

Read

paper/sections/*.tex

and build a slide-by-slide outline.

Slide template by talk type:

读取

paper/sections/*.tex

并逐张构建幻灯片大纲。

按报告类型划分的幻灯片模板：

Oral (15-22 slides)

口头报告（15-22张）

Slide	Purpose	Content Source	Figure?
1	Title	Paper metadata	No
2	Outline	Section headers	No
3-4	Motivation & Problem	Introduction	Optional
5	Key Insight	Introduction (contribution)	No
6-9	Method	Method section	Yes (hero figure)
10-14	Results	Experiments	Yes (per slide)
15-16	Analysis / Ablations	Experiments	Yes
17	Limitations	Conclusion	No
18	Conclusion / Takeaway	Conclusion	No
19	Thank You + QR	—	QR code

幻灯片序号	用途	内容来源	是否使用图片
1	标题页	论文元数据	否
2	报告大纲	章节标题	否
3-4	动机与问题阐述	引言部分	可选
5	核心创新点	引言（贡献部分）	否
6-9	方法介绍	方法章节	是（核心图）
10-14	实验结果	实验章节	是（每页一张）
15-16	分析/消融实验	实验章节	是
17	研究局限性	结论部分	否
18	结论/核心收获	结论部分	否
19	致谢+二维码	—	二维码

Spotlight (8-12 slides)

焦点报告（8-12张）

Slide	Purpose	Content Source	Figure?
1	Title	Paper metadata	No
2-3	Problem + Why It Matters	Introduction	Optional
4	Key Insight	Contribution	No
5-6	Method	Method (condensed)	Yes (hero)
7-9	Results	Key results only	Yes
10	Takeaway	Conclusion	No
11	Thank You + QR	—	QR code

幻灯片序号	用途	内容来源	是否使用图片
1	标题页	论文元数据	否
2-3	问题与重要性阐述	引言部分	可选
4	核心创新点	贡献部分	否
5-6	方法简介	方法章节（精简版）	是（核心图）
7-9	关键结果	仅展示核心实验结果	是
10	核心收获	结论部分	否
11	致谢+二维码	—	二维码

Poster-talk (5-8 slides)

海报闪电报告（5-8张）

Slide	Purpose	Content Source	Figure?
1	Title	Paper metadata	No
2	Problem	Introduction (1 slide)	No
3	Method	Method (1 slide)	Yes
4-5	Results	Key result only	Yes
6	Takeaway + QR	Conclusion	QR

For each slide, specify:

Title (max 8 words)
3-5 bullet points (max 8 words each)
Figure reference (if any) from paper/figures/
Speaker note (2-3 sentences of what to say)
Time allocation (in seconds)

Output:

slides/SLIDE_OUTLINE.md

🚦 Checkpoint:

📊 Slide outline ready:
- Talk type: [TALK_TYPE] ([TALK_MINUTES] min)
- Slide count: [N] slides
- Figures used: [N] from paper/figures/
- Time budget: [breakdown]

Slide-by-slide outline:
1. [Title slide]
2. [Motivation — 1.5 min]
3. [Problem statement — 1 min]
...

Proceed to drafting? Or adjust the outline?

⛔ STOP HERE and wait for user response. This is the most critical checkpoint — the outline determines the entire talk flow.

Options:

"go" → proceed to Phase 2
adjustments (e.g., "merge slides 3-4", "add a demo slide", "cut the ablation") → revise
"stop" → save to
```
slides/SLIDE_OUTLINE.md
```

State: Write

SLIDES_STATE.json

with

phase: 1

幻灯片序号	用途	内容来源	是否使用图片
1	标题页	论文元数据	否
2	问题阐述	引言（单页）	否
3	方法介绍	方法章节	是
4-5	核心结果	仅展示关键结果	是
6	核心收获+二维码	结论部分	二维码

每张幻灯片需明确：

标题（最多8个单词）
3-5个要点（每个要点最多8个单词）
图片引用（若有），来自paper/figures/目录
演讲者备注（2-3句话，提示演讲内容）
时间分配（秒为单位）

输出文件：

slides/SLIDE_OUTLINE.md

🚦 检查点：

📊 幻灯片大纲已生成：
- 报告类型：[TALK_TYPE]（[TALK_MINUTES]分钟）
- 幻灯片数量：[N]张
- 使用图片数量：[N]张，来自paper/figures/
- 时间分配明细：[breakdown]

逐张幻灯片大纲：
1. [标题页]
2. [动机 — 1.5分钟]
3. [问题陈述 — 1分钟]
...

是否继续进入草稿生成阶段？或需要调整大纲？

⛔ 在此处暂停，等待用户回复。 这是最关键的检查点——大纲决定了整个报告的叙事逻辑。

可选操作：

"go" → 进入阶段2
提出调整需求（例如："合并第3-4张幻灯片"、"添加演示幻灯片"、"移除消融实验部分"）→ 修改大纲
"stop" → 将大纲保存到
```
slides/SLIDE_OUTLINE.md
```

状态保存：写入

SLIDES_STATE.json

，标记

phase: 1

。

Phase 2: Slide-by-Slide Content Drafting

阶段2：逐张幻灯片内容撰写

For each slide in the outline, draft the actual content.

Presentation rules (enforced strictly):

Rule	Rationale
One message per slide	If a slide has two ideas, split it
Max 6 lines per slide	More than 6 lines = wall of text
Max 8 words per line	Audience reads, not listens, if text is long
Sentence fragments, not sentences	"Improves F1 by 3.2%" not "Our method improves the F1 score by 3.2 percentage points"
Figure slides: figure ≥60% area	The figure IS the content; bullets are annotations
Bold key numbers	"Achieves 94.3% accuracy"
Progressive disclosure	Use `\pause` or `\onslide` for complex slides
No Related Work slide	Unless invited talk (30+ min)

For each slide, produce:

```
\frametitle{}
```
Content (itemize or figure + caption)
```
\note{}
```
with speaker text (if SPEAKER_NOTES=true)

根据大纲逐张撰写幻灯片的实际内容。

严格执行以下演示规则：

规则	原因说明
每张幻灯片仅传递一个核心信息	若一张幻灯片包含两个独立观点，需拆分为两张幻灯片
每张幻灯片最多6行内容	超过6行会形成文字墙，影响听众理解
每行最多8个单词	文字过长会让听众专注阅读而非听演讲
使用短语而非完整句子	例如用“F1提升3.2%”代替“我们的方法将F1分数提升了3.2个百分点”
图片页：图片占比≥60%	图片是内容核心，要点仅作为注释
关键数据加粗	例如“准确率达94.3%”
逐步揭示复杂内容	对复杂幻灯片使用 `\pause` 或 `\onslide` 命令
除非是特邀报告，否则不包含相关工作幻灯片	普通报告时长有限，需聚焦核心内容

每张幻灯片需生成：

```
\frametitle{}
```
标题
内容（列表或图片+说明）
```
\note{}
```
演讲者备注（若SPEAKER_NOTES=true）

Phase 3: Generate Slides LaTeX

阶段3：生成Beamer LaTeX源码

Create

slides/main.tex

using beamer.

Template structure:

latex

\documentclass[aspectratio=169]{beamer}

% Venue theme
\usepackage{xcolor}
\definecolor{primary}{HTML}{VENUE_PRIMARY}
\definecolor{accent}{HTML}{VENUE_ACCENT}

% Clean theme
\usetheme{default}
\usecolortheme{default}
\setbeamercolor{frametitle}{fg=primary}
\setbeamercolor{title}{fg=primary}
\setbeamercolor{structure}{fg=accent}
\setbeamercolor{itemize item}{fg=primary}
\setbeamercolor{itemize subitem}{fg=accent}
\setbeamertemplate{navigation symbols}{}
\setbeamertemplate{footline}{
  \hfill\insertframenumber/\inserttotalframenumber\hspace{2mm}\vspace{2mm}
}

% Packages
\usepackage{graphicx,amsmath,booktabs}
\graphicspath{{figures/}}

% Speaker notes (if enabled)
% \setbeameroption{show notes on second screen=right}

% Metadata
\title{PAPER TITLE}
\author{Author 1 \and Author 2}
\institute{Affiliation}
\date{VENUE YEAR}

\begin{document}

\begin{frame}
\titlepage
\end{frame}

% Content slides follow...

\begin{frame}{Motivation}
\begin{itemize}
  \item Bullet point 1
  \item Bullet point 2
  \item \textbf{Key insight in bold}
\end{itemize}
\note{Speaker note: explain the motivation...}
\end{frame}

% Figure slide example
\begin{frame}{Method Overview}
\centering
\includegraphics[width=0.85\textwidth]{method_overview.pdf}
\vspace{0.5em}
\begin{itemize}
  \item Key annotation about the figure
\end{itemize}
\note{Walk through the figure left to right...}
\end{frame}

% ... more slides ...

\begin{frame}{Thank You}
\centering
{\Large Questions?}\\[2em]
Paper: [URL or QR placeholder]\\
Code: [URL or QR placeholder]
\end{frame}

\end{document}

Symlink figures:

bash

ln -sf ../paper/figures/*.pdf slides/figures/ 2>/dev/null
ln -sf ../paper/figures/*.png slides/figures/ 2>/dev/null

Key formatting rules:

Title font: ≥28pt, venue primary color
Body font: ≥20pt
Footnotes: ≥14pt
No navigation symbols
Frame numbers in bottom-right
Clean white background (no gradients, no decorative elements)

创建

slides/main.tex

文件，使用beamer格式。

模板结构：

latex

\documentclass[aspectratio=169]{beamer}

% 会议主题配色
\usepackage{xcolor}
\definecolor{primary}{HTML}{VENUE_PRIMARY}
\definecolor{accent}{HTML}{VENUE_ACCENT}

% 简洁主题
\usetheme{default}
\usecolortheme{default}
\setbeamercolor{frametitle}{fg=primary}
\setbeamercolor{title}{fg=primary}
\setbeamercolor{structure}{fg=accent}
\setbeamercolor{itemize item}{fg=primary}
\setbeamercolor{itemize subitem}{fg=accent}
\setbeamertemplate{navigation symbols}{}
\setbeamertemplate{footline}{
  \hfill\insertframenumber/\inserttotalframenumber\hspace{2mm}\vspace{2mm}
}

% 加载必要包
\usepackage{graphicx,amsmath,booktabs}
\graphicspath{{figures/}}

% 演讲者备注（若启用）
% \setbeameroption{show notes on second screen=right}

% 元数据
\title{PAPER TITLE}
\author{Author 1 \and Author 2}
\institute{Affiliation}
\date{VENUE YEAR}

\begin{document}

\begin{frame}
\titlepage
\end{frame}

% 内容幻灯片依次展开...

\begin{frame}{Motivation}
\begin{itemize}
  \item Bullet point 1
  \item Bullet point 2
  \item \textbf{Key insight in bold}
\end{itemize}
\note{Speaker note: explain the motivation...}
\end{frame}

% 图片页示例
\begin{frame}{Method Overview}
\centering
\includegraphics[width=0.85\textwidth]{method_overview.pdf}
\vspace{0.5em}
\begin{itemize}
  \item Key annotation about the figure
\end{itemize}
\note{Walk through the figure left to right...}
\end{frame}

% ... 更多幻灯片 ...

\begin{frame}{Thank You}
\centering
{\Large Questions?}\\[2em]
Paper: [URL or QR placeholder]\\
Code: [URL or QR placeholder]
\end{frame}

\end{document}

图片软链接：

bash

ln -sf ../paper/figures/*.pdf slides/figures/ 2>/dev/null
ln -sf ../paper/figures/*.png slides/figures/ 2>/dev/null

关键格式规则：

标题字体：≥28pt，使用会议主色调
正文字体：≥20pt
脚注字体：≥14pt
隐藏导航符号
幻灯片编号位于右下角
简洁白色背景（无渐变、无装饰元素）

Phase 4: Compile Slides

阶段4：编译幻灯片

bash

cd slides && latexmk -$ENGINE -interaction=nonstopmode main.tex

Error handling loop (max 3 attempts):

Parse error log
Fix: missing package, undefined command, file not found, overfull boxes
Recompile

Verification:

bash

undefined

bash

cd slides && latexmk -$ENGINE -interaction=nonstopmode main.tex

错误处理循环（最多3次尝试）：

解析错误日志
修复问题：缺失包、未定义命令、文件未找到、内容溢出
重新编译

验证步骤：

bash

undefined

Check slide count matches outline

检查幻灯片数量是否与大纲一致

pdfinfo slides/main.pdf | grep Pages


If page count differs significantly from outline (>2 slides off), investigate.

**State**: Write `SLIDES_STATE.json` with `phase: 4`.

pdfinfo slides/main.pdf | grep Pages


若实际页数与大纲差异较大（超过2张），需排查原因。

**状态保存**：写入`SLIDES_STATE.json`，标记`phase: 4`。

Phase 5: Codex MCP Review

阶段5：Codex MCP审核

Send the slide outline + selected LaTeX frames to GPT-5.4 xhigh:

mcp__codex__codex:
  config: {"model_reasoning_effort": "xhigh"}
  prompt: |
    Review this [TALK_TYPE] presentation ([TALK_MINUTES] min) for [VENUE].

    Evaluate using these criteria (score 1-5 each):

    1. **Story arc** — Does the talk build a compelling narrative? (Problem → insight → method → evidence → takeaway)
    2. **Slide density** — Any slides with too much text? (Max 6 lines, 8 words/line)
    3. **Time budget** — Is [N] slides realistic for [TALK_MINUTES] minutes?
    4. **Figure visibility** — Will figures be readable on a projector?
    5. **Opening hook** — Do slides 2-3 grab attention? (Not "In this paper, we...")
    6. **Takeaway** — Is the final message clear and memorable?
    7. **Progressive build** — Are complex ideas revealed gradually?

    Slide outline:
    [PASTE SLIDE_OUTLINE.md]

    Selected frames (LaTeX):
    [PASTE KEY FRAMES]

    Provide:
    - Score for each criterion
    - Top 3 actionable fixes
    - Overall: Ready to present? (Yes / Needs revision / Major issues)

Apply fixes. Recompile if LaTeX was changed.

⚠️ If
mcp__codex__codex
is not available (no OpenAI API key), skip external review and proceed to Phase 6. Note the skip in
SLIDES_STATE.json
.

Save review to

slides/SLIDES_REVIEW.md

State: Write

SLIDES_STATE.json

with

phase: 5

将幻灯片大纲+选中的LaTeX帧发送给GPT-5.4 xhigh模型：

mcp__codex__codex:
  config: {"model_reasoning_effort": "xhigh"}
  prompt: |
    请审核这份面向[VENUE]会议的[TALK_TYPE]报告幻灯片（时长[TALK_MINUTES]分钟）。

    请根据以下标准评分（1-5分）：

    1. **叙事逻辑** — 报告是否构建了引人入胜的完整叙事？（问题→创新点→方法→证据→核心收获）
    2. **幻灯片密度** — 是否存在文字过多的幻灯片？（最多6行，每行8个单词）
    3. **时间匹配度** — [N]张幻灯片是否适配[TALK_MINUTES]分钟的时长？
    4. **图片可读性** — 图片在投影仪上是否清晰可见？
    5. **开场吸引力** — 第2-3张幻灯片是否能抓住听众注意力？（避免以“在本文中，我们...”开头）
    6. **核心传递** — 最终结论是否清晰且令人印象深刻？
    7. **内容递进** — 复杂观点是否逐步揭示？

    幻灯片大纲：
    [粘贴SLIDE_OUTLINE.md内容]

    选中的LaTeX帧：
    [粘贴关键帧内容]

    请提供：
    - 各标准的评分
    - 最需改进的3个可执行建议
    - 整体结论：是否可直接用于报告？（是/需修改/存在重大问题）

根据建议修改内容，若LaTeX代码有变更则重新编译。

⚠️ 若
mcp__codex__codex
不可用（无OpenAI API密钥），跳过外部审核直接进入阶段6，并在
SLIDES_STATE.json
中标记跳过原因。

将审核结果保存到

slides/SLIDES_REVIEW.md

。

状态保存：写入

SLIDES_STATE.json

，标记

phase: 5

。

Phase 6: Speaker Notes

阶段6：演讲者备注完善

For each slide, ensure a

\note{}

block exists with:

What to say (2-3 complete sentences, conversational tone)
Timing hint (e.g., "spend 1 minute here", "quick — 20 seconds")
Transition phrase to the next slide (e.g., "So how do we actually implement this? Let me show you...")

Also generate

slides/speaker_notes.md

as a standalone backup:

markdown

undefined

确保每张幻灯片都有

\note{}

块，包含：

演讲内容提示（2-3句完整句子，口语化表达）
时间提示（例如：“此处花费1分钟”、“快速带过 — 20秒”）
转场语（例如：“那么我们具体如何实现呢？接下来为您展示...”）

同时生成独立备份文件

slides/speaker_notes.md

：

markdown

undefined

Speaker Notes

演讲者备注

Slide 1: Title

第1张：标题页

[No speaking — wait for introduction]

[无需发言 — 等待主持人介绍]

Slide 2: Motivation

第2张：动机

"Thank you. So let me start with the problem we're trying to solve..." [Time: 1.5 min]

“感谢主持人。接下来我将从我们要解决的问题开始讲起...” [时长：1.5分钟]

Slide 3: Problem Statement

第3张：问题陈述

"Specifically, the challenge is..." → Transition: "To address this, our key insight is..." [Time: 1 min]

...


**State**: Write `SLIDES_STATE.json` with `phase: 6`.

“具体来说，我们面临的挑战是...” → 转场语：“针对这一问题，我们的核心创新点是...” [时长：1分钟]

...


**状态保存**：写入`SLIDES_STATE.json`，标记`phase: 6`。

Phase 7: PowerPoint Export

阶段7：PPTX格式导出

Generate an editable PPTX using

python-pptx

bash

python3 -c "import pptx" 2>/dev/null || pip install python-pptx

Write

slides/generate_pptx.py

that:

Creates a PPTX with correct aspect ratio (16:9 → 13.33" x 7.5"; 4:3 → 10" x 7.5")
For each beamer frame:
- Creates a slide with matching layout
- Title in venue primary color, bold
- Bullet points with venue accent color markers
- Figures embedded as images (from slides/figures/)
- Speaker notes transferred to PPTX notes field
Title slide with special formatting (centered, larger title)
Thank You slide with centered text
Applies venue color scheme throughout

bash

cd slides && python3 generate_pptx.py

使用

python-pptx

生成可编辑的PPTX文件：

bash

python3 -c "import pptx" 2>/dev/null || pip install python-pptx

编写

slides/generate_pptx.py

脚本，实现以下功能：

创建符合宽高比的PPTX文件（16:9 → 13.33" x 7.5"; 4:3 → 10" x 7.5"）
对每个beamer帧：
- 创建匹配布局的幻灯片
- 标题使用会议主色调、加粗
- 要点使用会议强调色标记
- 嵌入slides/figures/目录中的图片
- 将演讲者备注同步到PPTX的备注栏
标题页使用特殊格式（居中、大字体）
致谢页使用居中文字
全程应用会议配色方案

bash

cd slides && python3 generate_pptx.py

Output: slides/presentation.pptx

输出文件：slides/presentation.pptx


> ⚠️ If `python-pptx` is not installed, skip with a note: "Install `pip install python-pptx` to enable PowerPoint export."

**State**: Write `SLIDES_STATE.json` with `phase: 7`.


> ⚠️ 若未安装`python-pptx`，则跳过此步骤并提示：“请安装`pip install python-pptx`以启用PowerPoint导出功能。”

**状态保存**：写入`SLIDES_STATE.json`，标记`phase: 7`。

Phase 8: Full Talk Script

阶段8：完整演讲稿生成

Generate

slides/TALK_SCRIPT.md

— a complete, word-for-word script for the talk.

This is different from speaker notes (brief reminders). The talk script is a full manuscript that can be read aloud or used for practice.

markdown

undefined

生成

slides/TALK_SCRIPT.md

— 报告的完整逐字稿。

逐字稿与演讲者备注（简短提示）不同，它是完整的文稿，可直接朗读或用于练习。

markdown

undefined

Talk Script: [Paper Title]

演讲稿：[论文标题]

Venue: [VENUE] [YEAR] Talk type: [TALK_TYPE] ([TALK_MINUTES] min) Total slides: [N]

会议：[VENUE] [年份] 报告类型：[TALK_TYPE]（[TALK_MINUTES]分钟） 总幻灯片数：[N]

Slide 1: Title [0:00 - 0:15]

第1张：标题页 [0:00 - 0:15]

[Wait for chair introduction]

"Thank you [chair name]. I'm [author] from [affiliation], and today I'll be talking about [short title]."

[等待主持人介绍]

“感谢[主持人姓名]。我是来自[机构]的[作者姓名]，今天我将介绍的研究是[简短标题]。”

Slide 2: Motivation [0:15 - 1:30]

第2张：动机 [0:15 - 1:30]

"Let me start with the problem. [Describe the real-world motivation in accessible terms]. This matters because [impact statement].

The current state of the art approaches this with [brief existing approach]. But there's a fundamental limitation: [gap statement]."

→ Transition: "So what's our key insight?"

“首先我想从我们要解决的问题讲起。[用通俗易懂的语言描述现实场景中的动机]。这项研究之所以重要，是因为[阐述影响]。

当前的主流方法通常是[简要介绍现有方法]。但它存在一个根本性的局限：[指出研究空白]。”

→ 转场语：“那么我们的核心创新点是什么呢？”

Slide 3: Key Insight [1:30 - 2:30]

第3张：核心创新点 [1:30 - 2:30]

"Our key observation is that [core insight in one sentence].

This leads us to propose [method name], which [one-sentence description]."

→ Transition: "Let me walk you through how this works."

“我们的核心发现是[用一句话概括核心创新]。

基于此，我们提出了[方法名称]，它[一句话描述方法]。”

→ 转场语：“接下来我将为您详细介绍它的工作原理。”

Slide 4-N: [Continue for each slide...]

第4-N张：[按每张幻灯片依次展开...]

...

Slide [N]: Thank You [TALK_MINUTES:00]

第[N]张：致谢 [TALK_MINUTES:00]

"To summarize: we've shown that [main result]. The key takeaway is [memorable final message].

The paper and code are available at the QR code on screen. I'm happy to take questions."

“总结一下：我们的研究证明了[核心结果]。最关键的收获是[令人印象深刻的最终结论]。

论文和代码的链接已展示在屏幕的二维码中，欢迎大家提问。”

Time Budget Summary

时间分配汇总

Slide	Topic	Duration	Cumulative
1	Title	0:15	0:15
2	Motivation	1:15	1:30
3	Key Insight	1:00	2:30
...	...	...	...
N	Thank You	0:15	[TALK_MINUTES]:00

Total: [sum] min (target: [TALK_MINUTES] min)

幻灯片序号	主题	时长	累计时长
1	标题页	0:15	0:15
2	动机	1:15	1:30
3	核心创新点	1:00	2:30
...	...	...	...
N	致谢	0:15	[TALK_MINUTES]:00

总时长：[sum]分钟（目标：[TALK_MINUTES]分钟）

Anticipated Q&A

预期问答环节

Q1: How does this compare to [strongest baseline]?

Q1：这项研究与[最强基线方法]相比如何？

A: "[Specific comparison with numbers]. Our advantage is particularly clear in [specific scenario], where we see [X%] improvement."

A：“[结合具体数据进行对比]。我们的方法在[特定场景]下优势尤为明显，性能提升了[X%]。”

Q2: What are the main limitations?

Q2：研究的主要局限性是什么？

A: "[Honest answer]. We see this as [future work direction]."

A：“[诚实回答局限性]。我们将此视为未来的研究方向：[阐述未来工作]。”

Q3: How computationally expensive is this?

Q3：该方法的计算成本如何？

A: "[Training/inference cost]. Compared to [baseline], our method requires [comparison]."

A：“[说明训练/推理成本]。与[基线方法]相比，我们的方法[成本对比]。”

Q4: Does this generalize to [related domain]?

Q4：该方法是否能推广到[相关领域]？

A: "[Answer based on paper's discussion section]."

A：“[基于论文讨论部分回答]。”

Q5: What's the most surprising finding?

Q5：研究中最令人惊讶的发现是什么？

A: "[Interesting insight from the experiments]."

A：“[分享实验中的有趣发现]。”

Q6: How sensitive is the method to [hyperparameter/design choice]?

Q6：方法对[超参数/设计选择]的敏感度如何？

A: "[Reference ablation study if available]."

A：“[参考消融实验结果回答]。”

Q7: What's the next step for this research?

Q7：这项研究的下一步计划是什么？

A: "[Future work from conclusion]."

A：“[基于结论中的未来工作回答]。”

Q8: [Domain-specific question]

Q8：[领域特定问题]

A: "[Answer]."

undefined

A：“[对应回答]。”

undefined

Final Output Summary

最终输出汇总

📊 Slide generation complete:
- Talk type: [TALK_TYPE] ([TALK_MINUTES] min) for [VENUE]
- Files:
  slides/
  ├── main.tex              # Beamer LaTeX source
  ├── main.pdf              # Compiled slides (primary output)
  ├── presentation.pptx     # Editable PowerPoint
  ├── SLIDE_OUTLINE.md      # Slide-by-slide outline
  ├── SLIDES_REVIEW.md      # GPT-5.4 review feedback
  ├── speaker_notes.md      # Per-slide speaker notes
  ├── TALK_SCRIPT.md        # Full word-for-word talk script + Q&A
  ├── SLIDES_STATE.json     # State persistence
  ├── generate_pptx.py      # PPTX generation script
  └── figures/              # Symlinked from paper/figures/

Next steps:
1. Practice with TALK_SCRIPT.md (read aloud, time yourself)
2. Edit presentation.pptx for visual tweaks (animations, custom graphics)
3. Review Anticipated Q&A section before the talk
4. Do a dry run with a colleague

State: Write

SLIDES_STATE.json

with

phase: 8, status: "completed"

📊 幻灯片生成完成：
- 报告类型：[TALK_TYPE]（[TALK_MINUTES]分钟），面向[VENUE]会议
- 生成文件：
  slides/
  ├── main.tex              # Beamer LaTeX源码
  ├── main.pdf              # 编译后的幻灯片（主要输出）
  ├── presentation.pptx     # 可编辑PowerPoint文件
  ├── SLIDE_OUTLINE.md      # 逐张幻灯片大纲
  ├── SLIDES_REVIEW.md      # GPT-5.4审核反馈
  ├── speaker_notes.md      # 逐张演讲者备注
  ├── TALK_SCRIPT.md        # 完整逐字演讲稿+问答预案
  ├── SLIDES_STATE.json     # 状态持久化文件
  ├── generate_pptx.py      # PPTX生成脚本
  └── figures/              # 从paper/figures/软链接的图片

后续建议：
1. 使用TALK_SCRIPT.md进行练习（大声朗读，计时）
2. 编辑presentation.pptx进行视觉优化（动画、自定义图形）
3. 报告前熟悉预期问答环节的内容
4. 与同事进行模拟报告演练

状态保存：写入

SLIDES_STATE.json

，标记

phase: 8, status: "completed"

。

Key Rules

核心规则

Large file handling: If the Write tool fails due to file size, immediately retry using Bash (
```
cat << 'EOF' > file
```
) to write in chunks. Do NOT ask the user for permission — just do it silently.
One message per slide. If a slide has two ideas, split it into two slides.
Do NOT fabricate data. All numbers must come from
```
paper/sections/*.tex
```
.
Bullet points only — never full sentences on slides. Sentence fragments are fine.
Figure slides: figure ≥60% of slide area. The figure IS the content.
Progressive disclosure: Use
```
\pause
```
or
```
\onslide
```
for complex method slides.
De-AI polish: Remove watch words from all slide text and talk script.
Do NOT hallucinate citations. Reference only papers cited in the paper.
Opening hook matters: Never start with "In this paper, we..." — start with the problem or a provocative question.
Font size minimums: Title ≥28pt, body ≥20pt, footnotes ≥14pt.
Feishu notifications are optional. If
```
~/.claude/feishu.json
```
exists, send notifications. If absent, skip.

大文件处理：若写入工具因文件大小失败，立即使用Bash命令（
```
cat << 'EOF' > file
```
）分块写入，无需询问用户，静默执行。
每张幻灯片仅传递一个核心信息：若一张幻灯片包含两个独立观点，必须拆分为两张。
不得编造数据：所有数据必须来自
```
paper/sections/*.tex
```
。
仅使用要点：幻灯片上禁止出现完整句子，短语即可。
图片页：图片占比≥60%：图片是内容的核心。
逐步揭示复杂内容：对复杂方法页使用
```
\pause
```
或
```
\onslide
```
命令。
去AI化润色：从所有幻灯片文本和演讲稿中移除AI生成痕迹明显的词汇。
不得虚构引用：仅引用论文中已有的文献。
开场要有吸引力：绝不能以“在本文中，我们...”开头，需从问题或场景切入。
字体大小下限：标题≥28pt，正文≥20pt，脚注≥14pt。
飞书通知可选：若存在
```
~/.claude/feishu.json
```
则发送通知，否则跳过。

Parameter Pass-Through

参数传递示例

/paper-slides "paper/" — talk_type: oral, venue: ICML, minutes: 20, aspect: 4:3, notes: false

Parameter	Default	Description
`venue`	NeurIPS	Conference for color scheme
`talk_type`	spotlight	oral/spotlight/poster-talk/invited
`minutes`	15	Talk duration
`aspect`	16:9	Aspect ratio (16:9 / 4:3)
`notes`	true	Generate speaker notes
`engine`	pdflatex	LaTeX engine
`auto proceed`	false	Skip checkpoints

/paper-slides "paper/" — talk_type: oral, venue: ICML, minutes: 20, aspect: 4:3, notes: false

参数名	默认值	说明
`venue`	NeurIPS	会议名称，决定配色方案
`talk_type`	spotlight	报告类型
`minutes`	15	报告时长（分钟）
`aspect`	16:9	幻灯片宽高比
`notes`	true	是否生成演讲者备注
`engine`	pdflatex	LaTeX引擎
`auto proceed`	false	是否跳过检查点直接执行