agentleFS
LambenthanLambenthan/empiricalwiki/.claude/skills/paper-draft/SKILL.md

paper-draft

Skill85 starsChanged 5 months agoMITView on GitHub

从 PAPER_PLAN 起草 LaTeX 论文:逐 section 从 wiki 取材撰写 + 生成 figures/tables + BibTeX 验证 + de-AI polish

Files
SKILL.md

/paper-draft

从 /paper-plan 生成的 PAPER_PLAN.md 起草完整 LaTeX 论文。 逐 section 撰写:每个 section 从 wiki 读取 claims/experiments/papers/concepts 取材, 生成 LaTeX + figures + tables。BibTeX 从 DBLP/CrossRef 获取(遵循 citation-verification), 完成后执行 de-AI polish(遵循 academic-writing)。 可选逐 section Review LLM review。输出可编译的 paper/ 目录。

Inputs

  • plan:PAPER_PLAN.md 的路径(如 wiki/outputs/paper-plan-sparse-lora-2026-04-08.md)
  • --review(可选):启用逐 section Review LLM review
  • --sections(可选):只写指定 section(如 --sections 3,4 只写 Method + Experiments),用于增量撰写

Outputs

  • paper/ 目录(在 wiki 项目根下):
    • paper/main.tex — 主文件(include 各 section)
    • paper/sections/introduction.tex
    • paper/sections/related_work.tex
    • paper/sections/method.tex
    • paper/sections/experiments.tex
    • paper/sections/conclusion.tex
    • paper/sections/appendix.tex(若有)
    • paper/figures/ — 生成的 figures(PDF/PNG)
    • paper/tables/ — 独立 table 文件(可选)
    • paper/math_commands.tex — 共享数学符号定义
    • paper/references.bib — 验证的 BibTeX 条目
  • wiki/log.md — 追加日志

Wiki Interaction

Reads

  • wiki/outputs/paper-plan-*.md — PAPER_PLAN(章节大纲、evidence map、figure plan、citation plan)
  • wiki/claims/*.md — target claims 的 Statement、Evidence summary、Conditions
  • wiki/experiments/*.md — Results、Analysis、key_result、metrics 数据
  • wiki/papers/*.md — Method、Results、Related(作为引用内容和 baseline 参考)
  • wiki/concepts/*.md — Definition、Formal notation、Variants(支持 Method 撰写)
  • wiki/topics/*.md — Overview、Timeline、SOTA tracker(支持 Introduction 上下文)
  • wiki/ideas/*.md — Motivation、Hypothesis(支持 Introduction 的 gap 叙述)
  • wiki/people/*.md — 人名和机构(引用格式)
  • wiki/graph/edges.jsonl — 关系图(构建论证逻辑链)
  • wiki/graph/open_questions.md — 已知局限(写 Limitations 和 Future Work)
  • .claude/skills/shared-references/academic-writing.md — 写作规范
  • .claude/skills/shared-references/citation-verification.md — 引用纪律

Writes

  • paper/ 目录(所有文件)
  • wiki/log.md — 追加操作日志

Graph edges created

  • 无(paper-plan 已创建 derived_from 边)

Workflow

前置:确认工作目录为 wiki 项目根(包含 wiki/、raw/、tools/ 的目录)。

Step 1: 初始化 Paper 目录

  1. 读取 PAPER_PLAN.md,提取 venue、title、section 列表
  2. 若 paper/ 目录已存在:
    • 备份为 paper.bak-{timestamp}/
    • 提示用户确认覆盖
  3. 创建目录结构:
    paper/
    ├── main.tex
    ├── math_commands.tex
    ├── references.bib
    ├── sections/
    └── figures/
  4. 从 templates/ 复制 venue 模板(若存在):
    • templates/{venue}.sty 或 templates/{venue}/
    • 若无模板:使用通用 article class,在 main.tex 中注明需要替换为正式模板
  5. 生成 math_commands.tex:
    • 从 wiki/concepts/ 中收集所有 Formal notation
    • 统一符号定义(向量、矩阵、集合、常用运算符)
  6. 生成 main.tex 骨架:
    \documentclass{article} % 替换为 venue template
    \input{math_commands}
    % packages
    \usepackage{booktabs,graphicx,amsmath,hyperref}
    
    \title{<title>}
    \author{} % 匿名提交时留空
    
    \begin{document}
    \maketitle
    \begin{abstract}
    % Step 3 生成
    \end{abstract}
    \input{sections/introduction}
    \input{sections/related_work}
    \input{sections/method}
    \input{sections/experiments}
    \input{sections/conclusion}
    \bibliography{references}
    \bibliographystyle{plain} % 替换为 venue 要求的 style
    % \input{sections/appendix} % 取消注释如需要
    \end{document}

Step 2: 生成 Figures 和 Tables

对 PAPER_PLAN 中 Figure Plan 的每个条目:

  1. Diagram 类型(架构图等):
    • 使用 TikZ 或 pgfplots 生成 LaTeX 原生图
    • 若过于复杂:生成 matplotlib Python 脚本 → 输出 PDF
    • 保存到 paper/figures/{figure-name}.pdf
  2. Plot 类型(实验结果图):
    • 从 wiki/experiments/{slug}.md 提取数据
    • 生成 matplotlib 脚本(遵循 academic-writing 的 figure 设计规范):
      • Colorblind-safe palette
      • Font size >= 8pt
      • Error bars / confidence bands
      • Clear legend
    • 执行脚本生成 PDF:
      python3 paper/figures/plot_{name}.py
    • 保存到 paper/figures/{figure-name}.pdf
  3. Table 类型:
    • 使用 booktabs 风格(toprule, midrule, bottomrule)
    • Best result bold, second-best underline
    • 直接嵌入 section .tex 文件(小表)或独立 paper/tables/{name}.tex(大表)

Step 3: 逐 Section 撰写

对每个 section(按 PAPER_PLAN 的大纲顺序),若 --sections 指定则只写指定 section:

3a. 收集素材

从 PAPER_PLAN 的 section 定义中提取:

  • 该 section 支撑的 claims
  • 对应的 wiki 页面列表
  • 计划的 figures/tables
  • 引用列表

读取所有相关 wiki 页面的对应 section:

  • Introduction → wiki/ideas/{idea}.md#Motivation + wiki/topics/{topic}.md#Overview
  • Related Work → wiki/papers/.md#Related + wiki/concepts/.md#Comparison
  • Method → wiki/concepts/*.md#Formal_notation + wiki/claims/{claim}.md#Statement
  • Experiments → wiki/experiments/.md#Results + wiki/experiments/.md#Analysis
  • Conclusion → wiki/graph/open_questions.md + wiki/claims/*.md#Open_questions

3b. 撰写 LaTeX

遵循 shared-references/academic-writing.md:

  • 按该 section 的 paragraph plan 撰写
  • 插入 \cite{key} 引用(key 从 citation plan 映射)
  • 插入 \ref{fig:name} / \ref{tab:name} 引用 figures/tables
  • 使用 math_commands.tex 中定义的符号
  • 每段以 topic sentence 开头
  • Experiments section:claim-first 结构("We claim X. To verify, we...")

3c. De-AI Polish

对每个撰写的 section 执行 de-AI polish(参照 academic-writing.md):

  1. 扫描并替换 AI 特征词汇(delve, leverage, utilize, comprehensive...)
  2. 移除 excessive hedging
  3. 变化 sentence openings(避免连续相同句式)
  4. 移除 filler sentences(不增加信息的句子)
  5. 确保 active voice 为主
  6. 检查 notation consistency

3d. 可选 Review LLM Review(--review)

若启用 --review,对每个 section:

mcp__llm-review__chat:
  system: "You are a senior ML researcher reviewing one section of a paper draft.
           Focus on: clarity, logical flow, claim-evidence alignment, notation consistency.
           Point out any remaining AI-sounding language patterns.
           Suggest specific rewrites for unclear passages."
  message: |
    ## Section: {section name}
    {LaTeX content}

    ## Claims this section should support
    {claims from matrix}

    ## Review this section for:
    1. Does it clearly support its target claims?
    2. Is the writing clear and precise?
    3. Any AI-generated language patterns remaining?
    4. Is the notation consistent with other sections?
    5. Missing content that reviewers will expect?

根据 Review LLM 反馈修改 section(行内修改,不重写整个 section)。

Step 4: 构建 Bibliography

遵循 shared-references/citation-verification.md:

  1. 收集所有 section 中使用的 \cite{key} 引用
  2. 对每个引用,从 PAPER_PLAN 的 citation plan 获取 BibTeX:
    • 已验证的:直接写入 references.bib
    • [UNCONFIRMED] 的:写入 references.bib 底部,带 % [UNCONFIRMED] 注释
  3. 排除未引用的条目(不使用 \nocite{*})
  4. 验证 BibTeX 格式正确性(每条有 title, author, year)
  5. 输出 bibliography 统计:
    references.bib: {N} entries, {M} verified, {K} [UNCONFIRMED]

Step 5: 全文 Cross-Review

完成所有 section 后:

mcp__llm-review__chat:
  system: "You are a senior ML researcher performing a final review of a complete paper draft.
           Focus on: cross-section coherence, claim-evidence thread (does the paper prove what it claims?),
           narrative flow, notation consistency across sections, figure/table referencing.
           This is NOT a line-by-line review — focus on structural and argumentative issues."
  message: |
    ## Full Paper Draft
    {concatenated LaTeX of all sections}

    ## Evidence Map
    {from PAPER_PLAN}

    ## Review Focus
    1. Does the paper tell a coherent story from Introduction to Conclusion?
    2. Are all claims from the matrix adequately supported in the text?
    3. Is there notation inconsistency between sections?
    4. Are all figures/tables referenced and discussed?
    5. Any redundancy between sections?
    6. Overall readiness for submission (1-10)?

根据 Review LLM 反馈做最终调整。

Step 6: 完成输出

  1. 确认所有文件已写入 paper/ 目录
  2. 验证基本完整性:
    • 所有 \input{sections/X} 的文件存在
    • 所有 \includegraphics{figures/X} 的文件存在
    • 所有 \cite{key} 在 references.bib 中有对应条目
    • 所有 \ref{label} 有对应 \label{label}
  3. 追加日志:
    python3 tools/research_wiki.py log wiki/ \
      "paper-draft | drafted {venue} paper '{title}' | {N} sections, {M} figures, {K} citations ({V} verified)"
  4. 输出到终端:
    # Paper Write Complete
    
    ## Files
    - paper/main.tex (master file)
    - paper/sections/ ({N} section files)
    - paper/figures/ ({M} figure files)
    - paper/references.bib ({K} entries, {V_count} [UNCONFIRMED])
    - paper/math_commands.tex
    
    ## Status
    - Sections written: {list}
    - De-AI polish: applied
    - Review LLM review: {yes/no, if yes: overall score}
    - [UNCONFIRMED] citations: {count} (resolve before /paper-compile)
    
    ## Next Steps
    - Run `/paper-compile paper/` to compile and verify
    - Resolve [UNCONFIRMED] citations manually
    - Run `/refine paper/main.tex --focus writing` for further polish

Constraints

  • 每个 section 从 wiki 取材:不凭空生成内容。每个技术 claim 必须追溯到 wiki 页面
  • BibTeX 遵循 citation-verification.md:从 DBLP/CrossRef/S2 获取,不从 LLM 记忆生成
  • de-AI polish 必选:每个 section 写完后必须执行 polish pass,不可跳过
  • figures 遵循 academic-writing.md:colorblind-safe、font size >= 8pt、vector format preferred
  • 匿名提交:不写作者名、机构、致谢(venue 匿名要求)
  • \nocite{*} 禁止:只引用实际使用的条目
  • notation 一致性:所有 section 使用 math_commands.tex 中的统一符号
  • 已有 paper/ 目录备份后覆盖:不直接覆盖,先备份
  • wikilink → \cite 转换:PAPER_PLAN 中的 [[slug]] 在 LaTeX 中转换为 \cite{key}
  • tables 使用 booktabs:不使用竖线和全网格

Error Handling

  • PAPER_PLAN 找不到:报错,建议先运行 /paper-plan
  • PAPER_PLAN 格式不完整:列出缺失 section,建议重新运行 /paper-plan
  • wiki 页面找不到(plan 引用的 claim/experiment/paper 不存在):警告并跳过该引用,标注缺失
  • figure 生成失败(matplotlib 错误):输出占位符 % TODO: generate figure {name},继续其他 section
  • BibTeX 全部获取失败:使用 [UNCONFIRMED] 占位,在终端报告需要手动处理的数量
  • Review LLM 不可用(--review 模式):跳过 section review 和 cross-review,标注「unreviewed」
  • venue template 不存在:使用通用 article class,在 main.tex 中注明
  • section 过长(超过 plan 的 page budget):警告,建议移至 appendix 或压缩

Dependencies

Tools(via Bash)

  • python3 tools/research_wiki.py log wiki/ "<message>" — 追加日志
  • python3 tools/fetch_s2.py search "<title>" — BibTeX fallback(S2 搜索)
  • python3 — 执行 matplotlib figure 脚本

MCP Servers

  • mcp__llm-review__chat — 逐 section review(可选,--review)+ 全文 cross-review(Step 5)

Claude Code Native

  • Read — 读取 wiki 页面和 PAPER_PLAN
  • Glob — 查找 wiki 页面
  • Write — 写入 paper/ 目录中的文件
  • Bash — 执行 figure 脚本、创建目录
  • WebFetch — DBLP / CrossRef BibTeX 获取

Shared References

  • .claude/skills/shared-references/academic-writing.md — 写作规范 + de-AI polish 规则 + figure 设计
  • .claude/skills/shared-references/citation-verification.md — BibTeX 获取流程 + [UNCONFIRMED] 协议

Called by

  • /research Stage 5(论文写作阶段)
  • 用户手动调用

Download now (ZIP)

  • Other versions5 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 4

    1. DAIR-AutoResearch/AutoSci★ 1.7k25% same
    2. PKU-DAIR/AutoSci★ 1.7k25% same
    3. skyllwt/AutoSci★ 1.7k25% same
    4. skyllwt/OmegaWiki★ 1.7k25% same

    Wrote their own version · 1

    1. cabbage2000-lab/paper-tutor-skills★ 43same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • Used alongsideProjects with it also use exp-pilot-eval, exp-pilot-run and 3 more.

    Other files people use alongside this one.

    Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
    • DAIR-AutoResearchSkill

      exp-pilot-eval

      Pilot result evaluation — read pilot results, apply success criteria, update idea page (pilot_result, failure_reason if failed), generate PILOT_VERDICT_REPORT. Called by /ideate Phase 5 after /exp-pilot-run.

      Also in 4 projects that use this

    • DAIR-AutoResearchSkill

      exp-pilot-run

      Pilot experiment execution — read Pilot Spec YAML, write pilot code, run experiment(Confirm with the user before operation and require the applicant to conduct manual inspection), return results. Called by /ideate Phase 5. Does NOT modify wiki pages or judge pass/fail.

      Also in 4 projects that use this

    • phanghonghaoSkill

      poster

      HTML/CSS A4 海报生成。从沙龙/活动 Markdown 文件生成科技风 A4 海报 PDF/PNG。Use when user mentions: poster, 海报, 海报生成, make poster, generate poster, 活动海报, 沙龙海报.

      Also in 4 projects that use this

    • MicrosoftSkill

      visualize

      Build interactive HTML explanations, charts, simulations, comparisons, and UI mockups. Use when a visual helps understanding; use artifact skills for Office deliverables.

      Also in 4 projects that use this

    • +1 more from DAIR-AutoResearch
  • In this repository30 other files it gives its agents.

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.