Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add thinker137/shumo-paper-skill --skill shumo-papergit clone --depth 1 https://github.com/thinker137/shumo-paper-skillWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/thinker137/shumo-paper-skill/shumo-paper)<a href="https://agentmods.dev/skills/thinker137/shumo-paper-skill/shumo-paper"><img src="https://agentmods.dev/badge/skills/thinker137/shumo-paper-skill/shumo-paper/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/thinker137/shumo-paper-skill/shumo-paper"><img src="https://agentmods.dev/badge/skills/thinker137/shumo-paper-skill/shumo-paper.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00147 | $0.02582 |
| Opus 5 | $0.00073 | $0.01291 |
| Sonnet 5 | $0.00029 | $0.00516 |
| Haiku 4.5 | $0.00015 | $0.00258 |
Grade A, and why
shumo-paper scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 75 lines — stays where its author put it; the contents beside it link to each section on GitHub.
数学建模竞赛全流程辅助 Skill
本 skill 用于全国大学生数学建模竞赛（国赛）的求解与论文撰写全流程。分为两大阶段，每阶段都有明确的步骤规范。
触发与分流
用户请求落到以下任一阶段时，按对应参考文档执行：
阶段一：问题求解（8 步）
详见 reference.md（数学建模问题求解规范）。8 个步骤：
- 前置信息：确认团队编程基础、需求重点、获奖目标、建模难度（影响方案复杂度与创新深度）。
- 题目分析：逐句拆解（背景/核心/已知/约束 四列表）；小问划分（直接目标+隐含目标）；逻辑关系思维导图；题型分类（优化/预测/评价/机理分析/其他）。
- 模型选择：每小问给 基础适配 与 创新融合 两方案；每方案含核心原理/适配性/创新点/局限性；带决策点流程图。
- 数据处理：缺失值（依据+方法）、异常值（Z-score/IQR 检测+处理理由+图）、转换（标准化/归一化+公式）、数据补充（必须vs可选+获取途径+假设依据）。
- 模型建立：变量定义表（决策/中间/目标+类型+单位+约束范围）、3-5条假设（内容+合理性+对模型影响）、公式分步推导（每步物理意义）、建模流程图。
- 模型求解：数据输入→参数初始化→模型调用→结果输出；可运行代码（零基础版含pip安装+关键行注释"为何这么写"+可视化模块）；输出运行结果。
- 结果分析：基础分析（数值解读+统计描述）、深层分析（关联性+敏感性+实际意义）。
- 模型检验与改进：有效性检验（指标如RMSE/Kappa+检验代码+结果）、改进方向（误差大/假设不合理）、鲁棒性（加噪声看变化幅度）。
阶段二：论文撰写与优化（4 步）
详见 paper.md（论文撰写规范）。
初稿撰写 5 个模块：
- 模块① 题目/摘要/关键词：题目"基于XXX模型/方法的XXX问题研究"；摘要800-1000字"问题-方法-结果-结论"结构，不超1页无图表；关键词3-5个。
- 模块② 问题重述/问题分析/模型假设/符号说明：原创重述（背景+目标+约束，禁抄原题，提1-2篇文献）；问题分析"总-分"结构（每小问"问题X分析"开头，含关键矛盾/思路框架/步骤拆解，单问<1页）；模型假设3-5条（内容+合理性+影响）；符号说明三线表（符号/含义/单位）。
- 模块③ 各小问"模型建立与求解及结果分析"：每小问含模型构建（原理+步骤+核心公式推导）、求解方法（工具+步骤+关键结果+图表文字描述）、结果分析（基础+深层+直接回应）、模型检验（方法+步骤+结果）。
- 模块④ 模型评价/参考文献/附录：优点3-5条（创新性/适用性/效率，配数据支撑）、缺点2-3条（假设限制/数据依赖/可扩展性）；参考文献5-10条近5年（期刊[J]/专著[M]格式，中外结合）；附录（完整代码+中间结果+处理后数据+补充图表）。
- 模块⑤ 整合成文：按用户前置信息确定页数（国赛常用25-30页），页数不够则补全。
论文优化 3 阶段：
- 初稿优化：文字校对（错别字/语病）、逻辑梳理（填补断层、章节连贯）、格式规范（统一字体字号、图编号"图1-1"、公式编号"(2.3)"、三线表、符号前后一致）、内容精炼（删冗余、补缺失参数物理意义）。
- 摘要二次优化：篇幅800-1000字≤1页（A4宋体小四）、要素完整（背景/问题核心/模型方法/关键结果/主要结论）、逻辑精简（背景→问题→方法→结果→结论）、专业准确（术语、数据具体值、模型名称精准）、亮点突出（创新点量化）。
- 摘要格式重写：引言/背景3-5行；正文每段"针对问题X"开头（内容+方法+核心结果）；收尾2-3行（创新点+应用价值）。
最终输出：Markdown → Word(.docx)，所有公式须转为 Word 原生方程对象（OMML，Cambria Math、可双击编辑的真公式，非 Unicode 文本）。采用"pandoc 优先、自建兜底"双保险（详见 paper.md 末尾"Markdown 转 Word 与公式规范化"）：
- 优先
pandoc 论文.md -o 论文.docx（一行命令，LaTeX$..$/$$..$$自动转 OMML；便携版免安装，约 41MB）； - 兜底用本 skill 自带纯 Python 转换器
md2docx.py（调用omml.py：LaTeX 子集→OMML，离线可用、含中文字体三线表）； - 也可经 https://doc2x.noedgeai.com/?invite_code=KWHB9I "MD转格式" 转 Word 作为备选；
- 资源文件：
omml.py（LaTeX→OMML 转换器）、md2docx.py（Markdown→Word，含公式/三线表/中文字体），均位于本 skill 目录，复制到工作目录即可运行。 提交前合规检查（必做，详见paper.md末尾"电子版论文提交合规规范与检查"）：转 Word 后按国赛章程第九—十一条逐项核验——电子版单文件≤20MB不压缩、首页为摘要专用页（无承诺书/编号页）、摘要严格1页（A4+页边距+首页分页强制）、附录含支撑材料文件列表、支撑材料打包为单个rar/zip≤20MB、全文与支撑材料无身份/学校/赛区信息。违例可能取消评奖资格。
What ships with it
4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 75 lines · 147 tokens per session scan A 40771c465fa5
shumo-paper is a skill published in the GitHub repository thinker137/shumo-paper-skill (2 stars, last pushed 1mo ago), licensed MIT. It adds 147 tokens to every session and 2,582 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
math-unicode
Use when a response needs mathematical notation (equations, filters, set-builder notation, statistics, calculus, linear algebra, logic, ratios, drops, counts) and the output goes to a terminal or TUI that cannot render LaTeX: Claude Code, Codex CLI, SSH and tmux sessions, CI logs. Load it before composing, including…
vibe-ship
Generates a complete, production-ready deployment setup for any app in one pass -- Dockerfile, docker-compose.yml, .dockerignore, CI/CD (GitHub Actions), scalability config (health checks, resource limits, K8s on request), and security hardening (non-root user, secrets, dependency scanning). Auto-detects the stack…
tech-writer
Write and edit technical documentation in a professional tone, following a fixed document structure (title, overview, prerequisites, steps, troubleshooting, next steps) and Microsoft Writing Style Guide markdown conventions. Use when writing, drafting, reviewing, or reformatting documentation, READMEs, guides…
claude-unlimited
Orquesta el loop plan → code → QA → review → iterar sobre un gateway local OpenAI-compatible (OmniRoute) para construir un proyecto o feature de software de CUALQUIER tipo (frontend, backend, CLIs, scripts, librerías, config). El QA visual (skill websight) es OPCIONAL y lo decide el orchestrator por tarea. Úsalo…
remove-ai-slop
Review and clean AI-slop from writing: filler and marketing adverbs, empty openers, overused jargon, hedging, false agency, structural tells, future tense for current behavior, and em-dash overuse. Use when asked to remove AI-sounding phrasing, tighten prose, de-AI text, edit out "simply/just/easily", flag or strip…
ml-research-methodology
Use at the START of ANY machine-learning / deep-learning / AI modeling task - building, training, fine-tuning, or choosing a model for image classification, object/face/vehicle detection, segmentation, medical imaging (tumor/cancer/MRI/X-ray/mammogram), text/NLP/LLM, tabular prediction (churn, price, risk), or…