Academic Paper Strategist
AAASS554/codex-academic-paper-skills
A skill your agent uses when the user needs to plan, de-risk, or ground a software engineering / computer science undergraduate thesis from a real codebase before final writing.
中文理工科研究生毕业论文的写作风格约束与审校。双模式：①写作模式——撰写/扩写正文时即按规范下笔 ②审校模式——对已写段落/章节做诊断与改写建议。核心目标是让论文读起来像活人写的毕业论文，不露 AI 痕迹，也不滑向过度口语。触发于用户请求撰写、润色、审校中文毕业论文正文、章节、摘要、引言、结论、方法说明、实验描述、图表题注等场景。
Install Thesis Polish
$ npx skills add lmcggg/graduate-thesis-polish-and-write-skill --skill thesis-polish -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install lmcggg/graduate-thesis-polish-and-write-skill thesis-polish --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/lmcggg/graduate-thesis-polish-and-write-skill.git skills-src && mkdir -p .claude/skills && cp -r skills-src/thesis-polish .claude/skills/thesis-polish && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "thesis-polish" agent skill from https://github.com/lmcggg/graduate-thesis-polish-and-write-skill/tree/main/thesis-polish into .claude/skills/thesis-polish/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "thesis-polish", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/lmcggg/graduate-thesis-polish-and-write-skill/tree/main/thesis-polishType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add lmcggg/graduate-thesis-polish-and-write-skill --skill thesis-polish -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install lmcggg/graduate-thesis-polish-and-write-skill thesis-polish --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/lmcggg/graduate-thesis-polish-and-write-skill.git skills-src && mkdir -p .agents/skills && cp -r skills-src/thesis-polish .agents/skills/thesis-polish && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "thesis-polish" agent skill from https://github.com/lmcggg/graduate-thesis-polish-and-write-skill/tree/main/thesis-polish into .agents/skills/thesis-polish/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "thesis-polish", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add lmcggg/graduate-thesis-polish-and-write-skill --skill thesis-polish -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install lmcggg/graduate-thesis-polish-and-write-skill thesis-polish --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/lmcggg/graduate-thesis-polish-and-write-skill.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/thesis-polish .cursor/skills/thesis-polish && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "thesis-polish" agent skill from https://github.com/lmcggg/graduate-thesis-polish-and-write-skill/tree/main/thesis-polish into .cursor/skills/thesis-polish/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "thesis-polish", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/lmcggg/graduate-thesis-polish-and-write-skill.git --path thesis-polish--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add lmcggg/graduate-thesis-polish-and-write-skill --skill thesis-polish -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install lmcggg/graduate-thesis-polish-and-write-skill thesis-polish --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/lmcggg/graduate-thesis-polish-and-write-skill.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/thesis-polish .gemini/skills/thesis-polish && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "thesis-polish" agent skill from https://github.com/lmcggg/graduate-thesis-polish-and-write-skill/tree/main/thesis-polish into .gemini/skills/thesis-polish/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "thesis-polish", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install lmcggg/graduate-thesis-polish-and-write-skill thesis-polishInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add lmcggg/graduate-thesis-polish-and-write-skill --skill thesis-polish -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/lmcggg/graduate-thesis-polish-and-write-skill.git skills-src && mkdir -p .github/skills && cp -r skills-src/thesis-polish .github/skills/thesis-polish && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "thesis-polish" agent skill from https://github.com/lmcggg/graduate-thesis-polish-and-write-skill/tree/main/thesis-polish into .github/skills/thesis-polish/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "thesis-polish", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add lmcggg/graduate-thesis-polish-and-write-skill --skill thesis-polish -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install lmcggg/graduate-thesis-polish-and-write-skill thesis-polish --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/lmcggg/graduate-thesis-polish-and-write-skill.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/thesis-polish .opencode/skills/thesis-polish && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "thesis-polish" agent skill from https://github.com/lmcggg/graduate-thesis-polish-and-write-skill/tree/main/thesis-polish into .opencode/skills/thesis-polish/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "thesis-polish", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
thesis-polish- GitHub stars
- 183
- Token cost
- ~1.9k tokens
- SKILL.md length
- 403 words
- Files
- 9
- Skills in repo
- 1
- Repo updated
- First seen
- Licence
- MIT
At a glance
中文理工科研究生毕业论文的写作风格约束与审校。双模式：①写作模式——撰写/扩写正文时即按规范下笔 ②审校模式——对已写段落/章节做诊断与改写建议。核心目标是让论文读起来像活人写的毕业论文，不露 AI 痕迹，也不滑向过度口语。触发于用户请求撰写、润色、审校中文毕业论文正文、章节、摘要、引言、结论、方法说明、实验描述、图表题注等场景。
- Works in 2 steps: 写作模式（style… → 审校模式（reviewer）：当用户把已写好的段落/章节/tex…
- Tasks that involve Essays and academic help
- SKILL.md covers 你在做什么, 两个核心自检问题（最高优先级）, token 成本控制（重要） and 范例论文机制, plus 5 more sections
- Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
What it does
Thesis Polish is an agent skill from lmcggg/graduate-thesis-polish-and-write-skill. 中文理工科研究生毕业论文的写作风格约束与审校。双模式：①写作模式——撰写/扩写正文时即按规范下笔 ②审校模式——对已写段落/章节做诊断与改写建议。核心目标是让论文读起来像活人写的毕业论文，不露 AI 痕迹，也不滑向过度口语。触发于用户请求撰写、润色、审校中文毕业论文正文、章节、摘要、引言、结论、方法说明、实验描述、图表题注等场景。
Its SKILL.md is about 1.9k tokens, which your agent loads only when the skill is triggered. The skill folder holds 8 other files (for example `README.md`, `meta-rules.md` and `patterns-layout.md`).
It sits in Education, covering Essays and academic help. The repository describes itself as: 中国理工科专业研究生毕业论文写作skill / graduate-thesis-polish skill. The licence is MIT.
When your agent uses it
- Tasks that involve Essays and academic help
Example prompts
- “/thesis-polish”
Requirements
- Pre-approved tools (allowed-tools): Read, Write, Edit, Grep, Glob, Bash
Workflow steps
2 steps, taken from the first numbered list in SKILL.md.
- 写作模式（style charter）：当用户请求撰写或扩写毕业论文的任何正文内容时，先加载本 skill 全部规则，按规则约束下笔，而不是写完再改。
- 审校模式（reviewer）：当用户把已写好的段落/章节/tex 文件交给你审，输出诊断报告或直接改写。
What it can do on your machine
Read from SKILL.md and the folder at commit dc35ec5. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves these tools, so the agent can use them without asking each time:
ReadWriteEditGrepGlobBash
From allowed-tools in the SKILL.md frontmatter.
Runs code
No scripts in the folder and no shell commands in SKILL.md (its code samples are markdown).
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Context cost
Thesis Polish loads about 1.9k tokens when it runs. Until then it costs about 45 tokens; SKILL.md has 403 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check: notes
The automated check noted patterns worth knowing about, such as sudo or a known installer.
- NotePre-approves every shell command (allowed-tools: Bash)SKILL.md
allowed-tools: Read, Write, Edit, Grep, Glob, Bash
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
SKILL.md
The full file from lmcggg/graduate-thesis-polish-and-write-skill at commit dc35ec5, republished under its MIT licence (© lmcggg). 403 words, ~1,866 tokens.
Download SKILL.mdSave it as
.claude/skills/thesis-polish/SKILL.md (or your agent's skills folder). This skill also uses 8 other files; get the full folder from GitHub.- name
- thesis-polish
- description
- 中文理工科研究生毕业论文的写作风格约束与审校。双模式：①写作模式——撰写/扩写正文时即按规范下笔 ②审校模式——对已写段落/章节做诊断与改写建议。核心目标是让论文读起来像活人写的毕业论文，不露 AI 痕迹，也不滑向过度口语。触发于用户请求撰写、润色、审校中文毕业论文正文、章节、摘要、引言、结论、方法说明、实验描述、图表题注等场景。
- allowed-tools
- Read, Write, Edit, Grep, Glob, Bash
- user-invocable
- true
thesis-polish: 中文理工科毕业论文写作风格规范
你在做什么
你是一个面向理工科研究生毕业论文的中文写作风格规范，覆盖工科、理科、信息、医工交叉等方向。职责有两层，同等重要：
- 写作模式（style charter）：当用户请求撰写或扩写毕业论文的任何正文内容时，先加载本 skill 全部规则，按规则约束下笔，而不是写完再改。
- 审校模式（reviewer）：当用户把已写好的段落/章节/tex 文件交给你审，输出诊断报告或直接改写。
用户的根本诉求是："一眼看上去就是活人写的中文毕业论文，不是 AI 堆的、不是翻译腔、不是学术会议论文。"
两个核心自检问题（最高优先级）
不论写作还是审校，每段输出前后都要用这两个问题过一遍：
- 删除测试：把这句话删掉、或换成更短的表达，意思有损失吗？没有损失就是 AI 填充，删。
- 范例测试：这句话/这个说法/这个小节名，有没有可能出现在用户指定的范例毕业论文里？不可能就是不合格。
这两个测试比任何具体黑名单都更本质。命中任一就修。
token 成本控制（重要）
调用 skill 时坚持以下原则，避免无谓浪费：
- 范例论文默认只读指纹：后续对话加载
reference-thesis-fingerprint.md（约 5-10 KB），不加载范例论文原文。指纹由 skill 在首次配置范例时一次性生成，详见"风格指纹"章节。 - 有更深需求时才回原文：指纹不足以判断时，先用 Grep 搜关键片段，再不够才 Read 原文某章（必须带 limit/pages）。绝不全量 Read 范例论文。
- patterns 文件按模块路由选加：根据当前任务场景（审摘要 / 审方法章 / 审实验章），只加载相关的 patterns-*.md，不全部加载。见下方"规则模块路由"表。
- 单次调用处理一个明确范围：避免"把整本论文审一遍"这种大任务；推荐按章或按节分批，输出更可读也更省 token。
范例论文机制
本 skill 强烈建议用户配置一本或多本范例论文（同方向、语言风格他认可的同类硕士/博士论文）。
启动流程
skill 被触发后第一件事：检查 reference-thesis.md 是否被用户填写。
- 已填写：只读取路径字段（不读范例论文的实际内容）。后续的"范例测试"具体化为"这话会出现在 {范例论文名} 里吗？"。范例论文本身按需访问，详见下面"按需访问"原则。
- 未填写，但用户本次给了路径：按下面"运行时声明"流程处理。
- 未填写且用户没给：向用户说明："本 skill 的效果依赖范例论文对照。你可以直接告诉我参考论文的路径（我会帮你记到 reference-thesis.md 里），或选择纯规则模式继续。" 等用户回复。
运行时声明（无需用户手动编辑 md）
当用户在对话里以任意方式声明范例论文时（例如"参考论文是 /path/to/xxx.pdf"、"范例就用 师兄的论文在 ~/docs/senior.tex"、"对照 A 和 B 这两本：…"），skill 自动完成以下动作：
- 提取信息：从用户消息里识别论文文件路径、论文名（若提到）、作者（若提到）。
- 验证路径：用 Bash
test -e {path}或 Read 首几行，确认文件存在且可读。路径不对直接问用户，不硬写进去。 - 写入 reference-thesis.md：用 Edit 把"我的范例论文"区块填上；信息不全时，作者/题目字段留空没关系，最关键的是文件路径。
- 生成风格指纹（见下）：一次性生成
reference-thesis-fingerprint.md。 - 回报用户：一句话确认写入成功（"已把范例论文设为 XXX，并提取了风格指纹，下次不用再说"），然后继续执行用户的主任务。
运行时更换 / 增补
- 用户说"换个参考论文"、"再加一本"、"不要那本了" 时，同样用 Edit 修改 reference-thesis.md 里的对应区块
- 多本参考时，第一本为"主要参照"，其余填入"次要参照"
- 范例论文变更时，同步重建
reference-thesis-fingerprint.md中对应的章节
风格指纹（核心：省 token 的关键机制）
范例论文本身通常几十万 token，不适合每次对话加载全文。skill 采用"风格指纹"机制：一次性从原文提取一个小文件 reference-thesis-fingerprint.md（约 5-10 KB），后续对话只读指纹，不再碰原文。
何时生成 / 更新
以下任一事件触发生成或更新指纹：
- 用户首次配置范例论文（运行时声明 或 手动填 reference-thesis.md）
- 用户更换 / 增补范例论文
- 用户显式要求"重建指纹"、"重新提取风格"
- 已有指纹文件过旧或缺失（即 reference-thesis.md 里的路径变了但指纹没更新）
指纹生成流程
按下面的步骤读原文，按指纹 schema 抽取内容并写入 ~/.claude/skills/thesis-polish/reference-thesis-fingerprint.md：
- Read 范例论文 → 如果是 PDF 先用
pages参数分批读；如果是 tex/txt 按章节读 - 定位关键章节：绪论/引言、方法章第一节、实验章第一节、任一章小结、任一小节级引言
- 按 schema 抽取内容（见下）
- 写入指纹文件；已有内容的按范例编号追加或替换
指纹 schema
每本参考论文占一个二级标题区块，内容包括：
## 范例 N: <论文名或路径文件名>
**元信息**
- 路径：<绝对路径>
- 方向/学科：<若知道>
- 指纹生成日期：<YYYY-MM-DD>
**章引言样例**（2-3 段原文节选，带章号）
> <原文段落 1>
> <原文段落 2>
**方法章叙述样例**（2 段原文节选）
> <原文段落>
**本章小结样例**（1-2 段原文节选）
> <原文段落>
**定理/证明叙述样例**（若有，1 段）
> <原文段落>
**图题样例**（5-10 条，标注图号）
- 图 X-Y: <图题>
**表题样例**（3-5 条）
- 表 X-Y: <表题>
**高频术语与句式**（10-20 条）
- <术语 1>
- <句式模板 1>（举一个具体例子）
**小节命名风格**（5-10 条）
- X.Y <小节名>
**公式叙述风格**（1 段节选，展示公式前后的文字衔接）
> <含公式的段落>
**指纹生成者备注**（可选，skill 写的一两句总结，比如"这本的语言偏向简洁、公式展开到物理量层、小结走三段式"）抽取时尽量用原文直引，不要复述——复述会丢风格信息。
分层访问策略
后续对话中遇到需要参考范例论文的情况，按以下阶梯：
- 默认（最省）：只加载
reference-thesis-fingerprint.md（约 5-10 KB）。绝大多数风格判断靠指纹完成。 - 需要对照具体表达时：用 Grep 在原文里搜 5-20 字的关键片段。Grep 只返回命中行，成本低。
- 指纹和 Grep 都不够（用户明确要求对照某一章风格、或有深度对齐需求）：Read 原文的某个章节，必须带
limit/offset或 PDF 的pages限定范围。 - 绝不：整本 Read、无边界 Read。
使用方式（遇到疑问时的流程）
当 skill 对某段文字不确定是否"像毕业论文"时：
- 先查本 skill 的各 patterns-*.md 黑名单
- 若仍不确定，查
reference-thesis-fingerprint.md里有没有类似场景的样例 - 仍不够，用 Grep 到范例原文里搜相似表达
- 再不够才 Read 范例的对应章节（带 limit/pages）
- 若范例里根本不会这样写，判定为不合格
工作流
入口判断
用户的请求大致分三类：
| 请求类型 | 标志 | 走哪个流程 |
|---|---|---|
| 要你写新内容 | "帮我写 XX 节"、"把这个思路扩成一段"、"写摘要" | A. 写作模式 |
| 要你审已有内容 | "检查这段"、"润色"、"看看哪里 AI 味重"、"审一下某章" | B. 审校模式 |
| 快速过硬规则 | "快速过一遍"、"只看硬伤" | C. 快速 checklist 模式 |
不明确时直接问用户走哪条。
Show full SKILL.md (157 more words)Show less
A. 写作模式
1. 加载 SKILL.md（已加载） + meta-rules.md（必读） + 按"模块路由"表只加载当前场景相关的 patterns
— 不要全部加载 patterns-*.md，会浪费 token
2. 检查 reference-thesis.md；未配置则按启动流程问用户
3. 生成草稿前，先在心里列出这段最可能踩到的 3-5 条规则，主动避开
4. 生成文本
5. 对输出做两遍自检：
a. 两个核心自检问题（删除测试 + 范例测试）
b. 按 meta-rules.md 的 K1/K2 过一遍（有没有把讨论内容写进正文？有没有 AI 痕迹词？）
6. 输出给用户
7. 不要在输出里加"我遵守了 XX 规则"之类的元评论——让文本说话关键：写作模式下，规则是生成时就要遵守的，不是写完再改。写完再改会留下痕迹，生成时避开才干净。
B. 审校模式
1. 加载 meta-rules.md + 按"模块路由"表只加载当前章节相关的 patterns
（如审摘要：patterns-vocabulary + patterns-syntax + patterns-layout + patterns-structure 的 I 类；不加载 patterns-math）
2. 读 reference-thesis.md 的路径字段（不读范例论文内容）
3. 读取用户给的范围（段落 / 整节 / 整个 tex 文件 / 多文件）
4. 先问用户默认行为：输出"诊断报告"还是"直接改写"？
（克制修改是默认立场——不确定时走诊断报告让用户决定）
5. 逐段扫描，对每处命中的规则记录：
- 定位（行号/段落号）
- 原文片段
- 命中的规则编号（如 A1、B2、D5）
- 严重程度：必改 / 建议 / 可选
- 改写建议（具体 before → after）
6. 仅对"不确定是否是 AI 味"的句子，用 Grep 到范例论文搜关键片段做对照（不 Read 全文）
7. 输出报告 / 改写后的文本
8. 终审两道自检（删除测试 + 范例测试）对 .tex 文件，尊重 LaTeX 指令：\ref、\cite、\begin{}..\end{} 等结构不要动，只改中文正文。公式内部（$...$、\begin{equation}...\end{equation}）不改。
C. 快速 checklist 模式
用本文件末尾的 20 条硬规则清单，对输入文本扫一遍，只报硬命中，不做软性建议。适合定稿前最后一遍过。
规则模块路由
写作或审校时按场景加载对应 patterns-*.md：
| 场景 | 必读 | 建议读 |
|---|---|---|
| 全文通用 | meta-rules.md, patterns-vocabulary.md, patterns-syntax.md, patterns-layout.md | - |
| 摘要 / 引言 | 上面 + patterns-structure.md (I1/I4/I5/I7) | - |
| 章小结 / 综合讨论 | 上面 + patterns-structure.md (I2/I3) | - |
| 方法章 | 上面 + patterns-math.md | patterns-structure.md (H) |
| 实验章 | 上面 + patterns-layout.md (F 图表类) | patterns-structure.md (D4) |
| 定理 / 证明 | 上面 + patterns-math.md (G4/J1/J2) | - |
| 目录 / 小节命名 | meta-rules.md + patterns-structure.md (D) | - |
具体模块：
- patterns-vocabulary.md — 词汇黑名单、术语恰当性
- patterns-syntax.md — 句式黑名单、长句控制
- patterns-layout.md — 标点、排版、图表
- patterns-structure.md — 小节命名、章引言、小结、摘要结构
- patterns-math.md — 公式叙述、Assumption/Remark/Theorem
- meta-rules.md — 讨论/正文分离、不暴露 AI、克制修改
- reference-thesis.md — 范例论文的配置与使用
输出规范
写作模式下
- 直接给论文正文，不加提示语、不加"根据规则 XX 写成"这类元评论
- 不用破折号、emoji、英文冒号
- 标点、术语、小节名都已经过自检
- 小标题不加粗
审校模式下（诊断报告形式）
## 审校结果
### 必改（X 条）
1. **【第 YY 行 / 第 Z 段】** 命中规则 B2（"X：Y"冒号拖尾句）
原文：<引用命中片段>
建议：<具体改法>
### 建议（X 条）
...
### 可选（X 条）
...审校模式下（直接改写形式）
- 保留原作者行文风格（克制原则）
- 只改命中规则的地方
- 句子原本通顺且无规则命中的，不动
- 末尾列一个简短的"本次修改命中的规则编号列表"，不要长篇自夸
20 条硬规则（快速 checklist）
- 不用破折号 ——，改逗号或拆句
- 不用"一是/二是"、"其一/其二"列举
- 不用"X：Y"冒号式拖尾句
- 不用"值得注意的是/不难看出/综上所述/本章将/首先...其次...最后..."
- 不用"进行+名词"堆砌（进行分析→分析）
- 不用"原因在于"，改"这是因为"
- 不先报框架再说内容（"本节从 X、Y、Z 三方面展开"禁用）
- 不在定理前写铺垫段
- 不加粗小节内强调、不给公式画外框、不用 emoji
- 图注简短，子图说明放正文，不写"实验一："前缀
- 小节标题不机械拼接、不用自造复合术语做标题
- 缩写第一次出现要写全称
- 摘要不堆缩写、不用中文冒号、不用破折号连路线
- 公式与正文之间保留空行
- 证明不"分三部分"、不(1)(2)、不加粗
- 长句 ≤60 字、相邻句不同开头、"的"字连用 ≤2 个
- 数字即结论，不前置"可以看出"、不后置"验证了有效性"
- 讨论内容绝不写入正文，不留 AI 痕迹
- 所有润色以用户指定的范例论文为风格基线，克制改写
- 章小结固定三段：提出什么 → 证明什么 → 实验结论（带数字），单段落
常见陷阱
- 过度去 AI 化变成"低于毕业论文水准"：去 AI 不是把话说短说浅。仍要保持严谨与专业（见 meta-rules.md K4）。
- 对话语写进正文：用户讨论时常用"我觉得"、"这个挺重要"，不能原样抄进正文。必须重新组织成独立学术表达（K1）。
- 为改而改：原文逻辑通顺、用词准确时，不要换同义词而重写（K5）。
- 自造术语显"精确"：研究者炫技思维是"越精准越好"，毕业论文思维是"越标准越好"。优先公认术语，避免自造复合名词作小节标题。
© lmcggg, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Files
SKILL.md and 8 other files in thesis-polish of lmcggg/graduate-thesis-polish-and-write-skill.
- SKILL.md
- README.md
- meta-rules.md
- patterns-layout.md
- patterns-math.md
- patterns-structure.md
- patterns-syntax.md
- patterns-vocabulary.md
- reference-thesis.md
Open the folder on GitHubat commit dc35ec5
Compare with similar skills
Thesis Polish next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Thesis Polish this skilllmcggg/graduate-thesis-polish-and-write-skill | 183 | — | ~1.9k | Automated safety check: Notes | MIT | |
| Academic Paper StrategistAAASS554/codex-academic-paper-skills | 539 | 1 repos | ~2.7k | Automated safety check: Pass | MIT | |
| Modeling Paper Rubric and Model Selectoryushui2022/MathModel-Skill | 454 | 1 repos | ~1.8k | Automated safety check: Pass | MIT | |
| Humanities Thesisganzhi-black/humanities-thesis-skill | 637 | — | ~1.7k | Automated safety check: Pass | MIT | |
| Skill Thesis Writeryanlin-cheng/skill-thesis-writer | 209 | — | ~1.6k | Automated safety check: Pass | Custom licence | |
| Thesis CreatorStars-OC/thesis-creator | 230 | — | ~2.8k | Automated safety check: Pass | MIT |
Similar skills
Modeling Paper Rubric and Model Selector
yushui2022/MathModel-Skill
Builds a scoring-aligned outline for a mathematical modeling paper and a model selection plan with baseline, improvement and validation experiments.
Humanities Thesis
ganzhi-black/humanities-thesis-skill
人文社科论文写作全流程指导。适用于文学、历史、哲学、社会学、传播学、新闻学、文化研究等领域的中文学术论文。当用户提到"论文""写论文""选题""文献综述""论文修改""论文结构""摘要翻译""帮我查文献""参考文献格式""论文没有新意""理论和文本脱节""章节之间缺乏逻辑""帮我检查论文""摘要翻译成英文""投稿准备""脚注格式"等场景时触发。覆盖从选题到投稿的全流程，包含防幻觉规则、学术数据库…
Skill Thesis Writer
yanlin-cheng/skill-thesis-writer
跨学科AI论文写作助手，专为本科生/研究生论文写作提供全方位支持。当用户需要撰写论文内容、设计论文框架结构、优化学术语体风格、处理参考文献格式（GB/T 7714-2015）、生成统计分析表格、或降低AI生成文本痕迹时使用此技能。支持工科（计算机/电子/机械等）、心理学、教育学、管理学等多学科领域，符合中国学术论文写作规范。
Thesis Creator
Stars-OC/thesis-creator
Walks Chinese undergraduates through writing a graduation thesis from topic to Word export, with text-similarity reduction and AI-text rate rewriting and checks.
Aigc Detector
free-revalution/AIGC-Detector-Pro
Academic paper AI content detection, rewriting, and thesis writing assistant.
Related
Categories
Questions about Thesis Polish
What does Thesis Polish do?
中文理工科研究生毕业论文的写作风格约束与审校。双模式：①写作模式——撰写/扩写正文时即按规范下笔 ②审校模式——对已写段落/章节做诊断与改写建议。核心目标是让论文读起来像活人写的毕业论文，不露 AI 痕迹，也不滑向过度口语。触发于用户请求撰写、润色、审校中文毕业论文正文、章节、摘要、引言、结论、方法说明、实验描述、图表题注等场景。. Thesis Polish is an agent skill from lmcggg/graduate-thesis-polish-and-write-skill.
When should I use Thesis Polish?
Thesis Polish fits situations like: tasks that involve Essays and academic help.
How do I install Thesis Polish in Claude Code?
Run `npx skills add lmcggg/graduate-thesis-polish-and-write-skill --skill thesis-polish -a claude-code`. Or copy the skill folder (thesis-polish in lmcggg/graduate-thesis-polish-and-write-skill) into .claude/skills/thesis-polish in your project. Claude Code loads it when a task matches its description.
How do I install Thesis Polish in Codex?
Run `npx skills add lmcggg/graduate-thesis-polish-and-write-skill --skill thesis-polish -a codex`. Or copy the skill folder (thesis-polish in lmcggg/graduate-thesis-polish-and-write-skill) into .agents/skills/thesis-polish in your project. Codex loads it when a task matches its description.
Can I use Thesis Polish in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add lmcggg/graduate-thesis-polish-and-write-skill --skill thesis-polish -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/thesis-polish, .gemini/skills/thesis-polish, .github/skills/thesis-polish and .opencode/skills/thesis-polish in your project.
What does Thesis Polish need to run?
SKILL.md names no scripts, command-line tools or credentials: Thesis Polish is instructions for the agent only. Its frontmatter pre-approves these tools: Read, Write, Edit, Grep, Glob, Bash.
Does Thesis Polish access the network?
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Is Thesis Polish safe to install?
Our automated static check of SKILL.md found notes only (pre-approves every shell command (allowed-tools: bash)), nothing it rates as a warning. It is not a guarantee. Review the folder before installing.
What licence does Thesis Polish use?
Thesis Polish is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does Thesis Polish use?
About 1.9k tokens (SKILL.md is roughly 7.5k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
What are the alternatives to Thesis Polish?
Skills that share tags, products or a category with Thesis Polish: Academic Paper Strategist (AAASS554/codex-academic-paper-skills, 539 stars), Modeling Paper Rubric and Model Selector (yushui2022/MathModel-Skill, 454 stars), Humanities Thesis (ganzhi-black/humanities-thesis-skill, 637 stars) and Skill Thesis Writer (yanlin-cheng/skill-thesis-writer, 209 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains Thesis Polish?
lmcggg (a GitHub user) maintains it in lmcggg/graduate-thesis-polish-and-write-skill, which has 183 GitHub stars. The repository was last updated on April 22, 2026.
Source: lmcggg/graduate-thesis-polish-and-write-skill on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.