No red flags found in any of the 11 categories — no credential harvesting, no data exfiltration, no curl-pipe-shell installer. Scanned against the SlowMist agent-security taxonomy, refreshed every 8 hours. Full audit →
by Lucasuiii · Claude Skill · ★ 77
Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h
🔒 Is cumcm-workflow safe to install? View the security audit →
CUMCM Workflow 简体中文 | English 一个给 AI agent 用的全国大学生数学建模竞赛工作流。它不替你想模型，它保证每一个写进论文的结论都能追回到官方题目和一次真实发生过的计算，并且在三天赛制下不把你拖垮。 支持 Codex 与 Claude Code。当前版本 v0.6，不向下兼容旧版本工作区。 快速开始 准备好 Codex 或 Claude Code、Python 3.10+，以及包含官方题目和附件的本地文件夹。 启动环境与路径： macOS / MacBook：打开有本地文件和终端访问能力的 Codex 或 Claude Code。资料路径示例：。 Windows（推荐 WSL2）：在 WSL2 的 Linux 环境中运行 agent 与工作流工具，由 agent 下载仓库。Windows 的 在 WSL 中通常对应 ；新项目路径可用 。Python、所选计算后端和 XeLaTeX 须在实际执行环境中可用，不能假定 Windows 上的安装会被 WSL 自动识别。 Windows 原生 PowerShell：文中的 Bash 示例（尤其 、、）需要改写，不能直接照抄。当前 CI 在 Linux 上运行，尚未完整验证 Windows 原生环境；WSL2 是推荐路径，并非已完成 Windows 端到端验证的声明。 提示词里的资料路径与输出路径必须使用 agent 实际运行环境的路径格式，不要混用 Windows 与 WSL 路径。 打开一个可写的本地工作目录，直接把下面的提示发给 AI，无需手动下载仓库或预先安装 Skill。将资料路径与输出路径替换成自己的绝对路径；输出目录应是尚不存在的新目录。AI 需要具备联网、读写本地文件和执行终端命令的权限。 已安装 Skill 时，也可用 Codex 的 或 Claude Code 的 触发；首次启动直接使用上述提示即可。 首次使用由 agent 检查所需依赖；计算阶段选择 MATLAB 或 Python，论文编译需要 XeLaTeX。你无需预先填写契约或运行每个脚
| Stars | 77 |
| Forks | 5 |
| Language | Python |
| Category | Claude Skill |
| License | MIT |
| Quality Score | 62.7881258910842/100 |
| Last Updated | 2026-09-19 |
| Created | 2026-08-29 |
| Platforms | claude-code, codex, python |
| Est. Tokens | ~17k |
Looking for a cumcm-workflow alternative? If you're comparing cumcm-workflow with other claude skill tools, these 6 projects are the closest alternatives on Agent Skills Hub — ranked by topic overlap, star count, and community traction.
数学建模竞赛完整工具链：从拿到赛题到交出论文，一条龙解决。 覆盖 国赛 CUMCM（A/B/C） 和 美赛 MCM/ICM（A-F） 全部题型。
三竞赛 (CUMCM/MCM/电工杯) 数学建模 skill — harness-agnostic, 同时支持 Claude Code 与 Codex CLI, 全程问答式 (Friendly Mode), 10 阶段
Frontend design skill for Claude Code, Codex, and Gemini CLI. Eight aesthetic anchors, each locking palette, t
Complete Kaggle integration plugin/skill for AI coding agents — competition reports, dataset/model downloads,
DeepContext is an MCP server that adds symbol-aware semantic search to Claude Code, Codex CLI, and other agent
World-class creative director in a skill. Methodology-driven ideation (SIT, TRIZ, Lateral Thinking, bisociatio
Explore other popular claude skill tools:
cumcm-workflow is Contest-native, evidence-focused CUMCM workflow for Codex and Claude Code. It is categorized as a Claude Skill with 77 GitHub stars.
cumcm-workflow is primarily written in Python. It covers topics such as agent-skill, claude-code, codex.
You can find installation instructions and usage details in the cumcm-workflow GitHub repository at github.com/Lucasuiii/cumcm-workflow. The project has 77 stars and 5 forks, indicating an active community.
cumcm-workflow is released under the MIT license, making it free to use and modify according to the license terms.
The top alternatives to cumcm-workflow on Agent Skills Hub include math-modeling-skills, mathmodel-skill, frontend-design. Each offers a different approach to the same problem space — compare them side-by-side by stars, quality score, and community activity.
Grades come from a rule-based scan built on the SlowMist agent-security taxonomy, covering 11 red-flag categories including credential harvesting, data exfiltration, and curl | sh installers. It is a first-layer scan, not a manual audit — we say so rather than overstate it.
The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.
Sources & who's responsible: