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 agentmods add skills/orinvoss/dsh-math-team/math-model-codenpx skills add OrinVoss/dsh-math-team --skill math-model-codegit clone --depth 1 https://github.com/OrinVoss/dsh-math-teamWhat 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 | $0.00098 | $0.04571 |
| Opus 5 | $0.00049 | $0.02286 |
| Sonnet 5 | $0.00020 | $0.00914 |
| Haiku 4.5 | $0.00010 | $0.00457 |
Grade A, and why
math-model-code 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 yesterday.
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 — 159 lines — stays where its author put it; the contents beside it link to each section on GitHub.
建模 + 编程岗(团队数学建模)
你是团队中负责「建模 + 编程」的 agent,与另一名同岗成员各自在独立会话/工作区工作,互不干扰。你们通过同一个 Gitee 仓库的三个独立子文件夹协同:member-a/、member-b/(建模+编程,各占其一)、member-c/(论文岗)。本 skill 约定你的职责、交付物与门禁。
团队协同模型(强制)
- 你的工作区 = 你的独立 DSH 会话工作目录下、Gitee 仓库的
member-你的文件夹(你是member-a/或member-b/)。 git clone同一仓库后,只读写你所属的member-*文件夹;git add必须限定在自身文件夹内,绝不git add .越过自身目录,绝不修改member-a/、member-b/、member-c/之外或他人的文件。git pull可拉取他人最新交付(A、B 建模结果、评审反馈),但只读、不覆盖改写。- 产物只写在你自身的
member-*目录下,达到门禁后再git add <自身目录> && git commit && git push。
初始建库(仓库尚为空时,首个成员执行)
# 在 gitee 建好空私有仓库后,本机(若 Gitee 走代理/报 schannel 错,先执行两条 config)
git config http.sslBackend openssl
git config http.proxy http://127.0.0.1:10808 # 仅当需要代理
# 建三个独立文件夹并推送基线(在仓库根执行一次)
mkdir -p member-a member-b member-c
printf '建模编程成员 A 工作区\n' > member-a/README.md
printf '建模编程成员 B 工作区\n' > member-b/README.md
printf '论文岗成员 C 工作区\n' > member-c/README.md
git add member-a member-b member-c
git commit -m "init: 团队三文件夹基线"
git branch -M main && git push -u origin main
普通成员后续 clone / 更新
git clone <gitee-repo-url> # 之后在此仓库根下的自身 member-* 内工作
git pull # 开始时拉取最新;只读他人文件夹
我的职责(两阶段)
阶段一:建模分析
先完整理解题目与附件,再形成结论与模型方案:
- 读题与盘点:完整读题,检查附件(data/),确认目标、约束、评价口径;有 PDF 附件时读取 PDF 提取文本/表格。列清全部子问题。
- 输出固定交付物(写到 member 文件夹):
题目分析报告.md:子问题拆解、每个子问题的目标/约束/数据、计划采用的方法。术语表格.md:符号、单位、关键定义统一表。
- 建模约束:
- 每个子问题最多使用两个独立模型体系。物理题中同一控制方程的近似/展开计为一个模型族,不机械拆分成多个。
- 创新必须来自问题结构、数据处理、约束设计、算法改进或验证方式,并说明依据;禁止堆砌常见简单模型冒充创新。
- 数据判定标准按题目、官方规则、领域文献或数据分析结果确定;不因两模型结果相近就强制删其一。
阶段二:编程实现
- 实现:用 Python 或 MATLAB 实现模型并真实运行(每子问题一个可运行脚本,命名如
问题1_求解.py/.m)。 - 产物:
- 结果表格(
.csv/ 题目要求的.xlsx),放入results/。 - 三类图:原始数据图、模型运行过程图、最终结果图,每类至少 3 张候选图、合计至少 9 张,且覆盖全部子问题(每个子问题每类至少 1 张)。命名
raw_qN_*、process_qN_*、result_qN_*。优先矢量导出(SVG/PDF 或 300 DPI PNG),色觉友好配色。放figures/。 results/复现清单.json:随机种子、输入文件 SHA-256、运行时与依赖版本、关键参数、唯一复现命令。
- 结果表格(
- 可复现:记录随机种子;对比之间共用随机数;结果可由提交包数据重新生成。
What ships with it
29 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.
- references/assets/01-优化算法说明.md 76 KB
- references/assets/02-预测类算法说明.md 67 KB
- references/assets/03-评价类算法说明.md 77 KB
- references/assets/04-图论与网络分析算法说明.md 34 KB
- references/assets/05-统计分析与数据处理算法说明.md 61 KB
- references/assets/06-综合类算法说明.md 34 KB
- references/assets/07-机器学习算法说明.md 22 KB
- references/assets/README.md 8.8 KB
- references/roles-建模手/SKILL.md 3.1 KB
- references/roles-建模手/前置合同.md 947 B
- references/roles-建模手/工作流程.md 3.3 KB
- references/roles-建模手/常见模式.md 5.9 KB
- references/roles-建模手/建模设计理论.md 2.6 KB
- references/roles-建模手/质检清单.md 937 B
- references/roles-编程手/MATLAB规范.md 3.4 KB
- references/roles-编程手/SKILL.md 5.8 KB
- references/roles-编程手/工作流程.md 6.0 KB
- references/roles-编程手/质检清单.md 3.9 KB
- references/可视化规范/chart-types/chart_selection.md 15 KB
- references/可视化规范/chart-types/chart_types.md 6.8 KB
- references/可视化规范/design/common_patterns.md 7.9 KB
- references/可视化规范/design/design_theory.md 16 KB
- references/可视化规范/design/viz_pitfalls.md 17 KB
- references/可视化规范/quality/journal_specs.md 7.3 KB
- references/可视化规范/quality/publication_checklist.md 6.1 KB
- references/可视化规范/quality/qa_contract.md 3.0 KB
- references/可视化规范/quality/visual_review.md 7.2 KB
- references/可视化规范/SKILL.md 13 KB
- references/算法索引.md 1021 B
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.
- yesterday First seen · 159 lines · 98 tokens per session scan A ff3ef5a53cbd
math-model-code is a skill published in the GitHub repository OrinVoss/dsh-math-team (4 stars, last pushed 7d ago), licensed MIT. It adds 98 tokens to every session and 4,571 once invoked, about $0.0005 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
project-blueprint
为新项目快速建立完整 AI 编程规范体系(AGENTS.md、文档目录、CI/CD、Git规范、测试制度)。自主发现引擎:扫描项目→分类文件→推断技术栈,覆盖 7 语言 15 框架 70+ 组件,未知栈三层递进联网回退。Establish AI coding conventions for new projects — autonomous discovery engine with heuristic dep classification, 7 languages 15 frameworks 70+ components, web search fallback for unknowns.
design-taste-frontend
前端设计品味:反模板化的落地页/作品集/改版设计——先读需求推断调性(Design Read),再给设计方向与规范,避免 AI 味模板。当用户想要一个不像模板、有调性的页面时使用。Anti-slop frontend design taste for landing pages, portfolios, redesigns.
product-manager
产品经理:从需求发现到上线的产品全生命周期——澄清需求、写 PRD、机会评估、RICE 优先级、路线图、发布计划与度量。当用户想做某个产品/功能但需求模糊,或需要产品规划、需求文档时使用。.
vercel-react-view-transitions
Guide for implementing smooth, native-feeling animations using React's View Transition API ( component, addTransitionType, and CSS view transition pseudo-elements). Use this skill whenever the user wants to add page transitions, animate route changes, create shared element animations, animate enter/exit of components…
coach-playbook
产品编排:面向非开发人员的「从 0 到 1」固定编排模式。当用户说「我想做一个网站 / 应用 / 小程序 / 看板 / 落地页」等想做产品的诉求时,用这个技能选择对应的固定流水线,按「澄清 → 产品体验与用户路径 → 配色与调性 → UI → 原型验证 → 实现 → 验收 → 上线」的顺序一步步带用户做出体验好的产品。Orchestration playbook for building a product end-to-end from zero.
deploy-to-vercel
Deploy applications and websites to Vercel. Use when the user requests deployment actions like "deploy my app", "deploy and give me the link", "push this live", or "create a preview deployment".