MathModeling-skills is a collection of Claude Code and Codex skills for mathematical modeling competitions, guiding users through data analysis, method-specific checks, reproducible computations, figures, and paper audits without taking over modeling decisions. It is for contestants who want AI to handle mechanical checks while they retain responsibility for choosing methods and interpreting results. The catalogue entries form the project's skills, instructions, hook, setting, and plugin workflow.
Borrowing it
Nothing to install: this file belongs to zhnnky329/MathModeling-skills. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/zhnnky329/MathModeling-skills/main/.claude/skills/decision-prompt-builder/SKILL.mdgit clone --depth 1 https://github.com/zhnnky329/MathModeling-skillsWrote 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/zhnnky329/mathmodeling-skills/decision-prompt-builder)<a href="https://agentmods.dev/skills/zhnnky329/mathmodeling-skills/decision-prompt-builder"><img src="https://agentmods.dev/badge/skills/zhnnky329/mathmodeling-skills/decision-prompt-builder/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/zhnnky329/mathmodeling-skills/decision-prompt-builder"><img src="https://agentmods.dev/badge/skills/zhnnky329/mathmodeling-skills/decision-prompt-builder.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 2 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Excessive Agency · line 44 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
- medium Excessive Agency · line 95 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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 | $0.00049 | $0.00754 |
| Opus 5.5 | $0.00020 | $0.00302 |
| Sonnet 5.5 | $0.00010 | $0.00151 |
| Haiku 4.5 | $0.00005 | $0.00075 |
Grade A, and why
decision-prompt-builder 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 2mo 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 — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Purpose
Ask the smallest useful question that only the human modeler can answer. Present mutually exclusive options with consequences; do not turn mechanical checks into user questions.
Inputs
- Current gate and the judgment it needs.
- Problem goal, required output, hard constraints, and available evidence.
planning/session_config.json.- Existing decisions in
methods/Qx/qx_decisions.jsonl.
Configuration
- Read
interaction_mode; accept legacymodefor compatibility. learning: show 2–3 short questions and withhold the AI suggestion until the user answers.speed: show one compressed question and optionally show the AI suggestion alongside.rigor_profiledoes not change who owns the judgment.
Choice-Card Workflow
- Identify one load-bearing judgment.
- Create 2–3 mutually exclusive options. Each option must state its practical consequence.
- Add
都不合适 / 补充约束when the listed options may not cover the user's intent. - Ask no more than three questions in one card.
- Do not recommend an option in
learningmode before the answer. - Pass the answer verbatim to
modeler-decision-logger; do not create a per-skill pending decision file.
Standard Cards
Before method screening
Ask only the missing high-impact items:
- output form to defend;
- interpretability/performance priority;
- unacceptable failure;
- experiment budget.
Do not ask the user to choose an algorithm name before evidence exists.
Example:
请选择这轮方案的首要取向：
- A. 可解释性优先——方法更透明，但可能牺牲部分拟合效果。
- B. 平衡——接受中等复杂度，要求能解释且优于可信 baseline。
- C. 性能优先——允许更复杂的方法，但需要额外稳健性和解释工作。
- D. 都不合适 / 我补充约束。
After a meaningful experiment
Use computed evidence to ask:
- proceed with the current main method;
- adjust a stated assumption or parameter and rerun;
- activate the recorded fallback.
Name the consequence and evidence for each option. Do not silently convert an AI metric preference into the human verdict.
Before final freeze
Use only when claim scope or confidence is genuinely judgment-bearing:
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.
- 2mo ago First seen · 103 lines · 49 tokens per session scan A 94c2b80edd3c
decision-prompt-builder is a skill published in the GitHub repository zhnnky329/MathModeling-skills (1,080 stars, last pushed 1mo ago), licensed MIT. It adds 49 tokens to every session and 754 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-08-30.
decision-prompt-builder compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| decision-prompt-builderzhnnky329/MathModeling-skills | Skill | 1.1k | 1mo ago | MIT | A | 49 | Claude Code |
| imagegenopenai/codex | Skill | 133k | yesterday | Apache-2.0 | A | 113 | Codex |
| playwright-climicrosoft/playwright | Skill | 99k | yesterday | Apache-2.0 | A | 19 | Claude Code |
| local-ai-agentsmicrosoft/ai-agents-for-beginners | Skill | 79k | 4d ago | MIT | A | 200 | Claude Code, Codex |
| systematic-debuggingobra/superpowers | Skill | 313k | 3d ago | MIT | A | 21 | Claude Code |
| next-cache-components-adoptionvercel/next.js | Skill | 143k | yesterday | MIT | A | 95 | Claude Code |
| next-partial-prefetching-adoptionvercel/next.js | Skill | 143k | yesterday | MIT | A | 103 | Claude Code |