science by brycewang-stanford
数值实验与算例分析（cjms-numerical-experiments） brycewang-stanford/cjms-numerical-experiments
Use when designing the numerical experiments and 算例 of a 《中国管理科学》 (Chinese Journal of Management Science) manuscript — grounded parameter calibration, an experiment matrix, sensitivity analysis, and mechanism-level reading of results. For simulation and case computation; real-data method validation belongs to cjms-empirical-validation.
What it changes View full report → Passed 20 of 22 test cases with the skill, and 7 of 22 without it.
Without skill 32%
With skill 91%
Checks it passed that the model alone missed and 22 more in the full report
13 fixed7 passed either way2 still failing0 broke
Pass-rate lift +59 pp
Tokens +29%
Agent turns 0%
Response time -10% Model gemini-3.6-flashMeasured 8/31/2026 v1 20 of 22 cases passed with skill
Use this skill Claude Code DecimalAI agent CLI & files
Save it into your project's skills folder. Claude Code picks it up in its next session.
curl -fsSL --create-dirs -o .claude/skills/brycewang-stanford-cjms-numerical-experiments/SKILL.md https://app.decimal.ai/s/brycewang-stanford-cjms-numerical-experiments/SKILL.md Copy command Copy SKILL.md
Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add brycewang-stanford-cjms-numerical-experiments, and start Claude Code with claude --plugin-dir "$(git rev-parse --show-toplevel)/.agents"
Linked stays up to date: your agents get the author's new versions automatically.
Editable copy is yours to change; the author's updates won't reach it.
Install linked skill Create editable copy
1 · One-time — install the CLI (
docs ↗ )
$ pip install "decimalai>=0.15"Copy
2 · Get the skill — writes .agents/skills/brycewang-stanford-cjms-numerical-experiments/ and pins it in .decimal/skills.lock
$ decimalai skills add brycewang-stanford-cjms-numerical-experimentsCopy
Copy link Copy README badge Copy SKILL.md No CLI? Get the same file — curl · raw file $ curl -fsSL --create-dirs -o .claude/skills/brycewang-stanford-cjms-numerical-experiments/SKILL.md https://app.decimal.ai/s/brycewang-stanford-cjms-numerical-experiments/SKILL.mdCopy
Safety checks: 3 of 3 passed
Code scan Clean Intent Passed Content Clean
Overview Benchmark Contents Activity Versions 1 Ratings 0
See it in action One of 13 test cases the skill fixed. Same model, same prompt, with and without the skill.
Prompt 我的算法在生产成本区间 [10, 50] 内都显著优于传统 heuristic 算法。如果继续增大成本到 200，新算法的优势可能会减弱甚至被超越。我应该把测试范围限制在 [10, 50] 还是继续往大扫？
What was checked, without and with the skill Checked Without skill With skill Advises sweeping parameters beyond the advantageous range to identify where advantage reverses. Demands reporting the method reversal region under 【边界发现】.
Show both answers
What's inside The instructions your model reads when the skill loads.
SKILL.md 90 lines single file
数值实验与算例分析（cjms-numerical-experiments）
触发时机
只有一组参数、一张图，审稿说"数值实验单薄"
参数取值说不出来源，被问"为什么 λ=0.5"
结果只描述曲线走向，没有机制解释与管理含义
核心：算例是论证，不是演示
本刊的应用建模传统里，算例承担双重责任 ：验证方法性质（收敛、优势、边界）+ 把抽象结论翻译回管理情境。一组"跑通了"的数字不构成算例分析。
参数标定的三个合法来源
拍脑袋参数只允许出现在"结构性质与取值无关"已被证明的场合，且需声明。教学示意数字必须标注"示意" ，不得伪装成标定值。
实验矩阵设计
按"要回答的问题"组织实验，而非按"能画的图"：
有效性 ：新方法 vs 基线（衔接 cjms-solution-algorithm 的对比设计），报解质量与时间。敏感性 ：核心参数逐一扫区间，其余固定在基准值；关键交互参数做双向网格。边界 ：找到方法失效或优势逆转的参数区域——审稿人最信任报告自身边界的论文。情境还原 ：至少一个贴近真实规模/取值的算例，支撑管理启示。
随机实验固定种子并报告重复次数与均值±标准差；实验环境（CPU、内存、软件版本）在脚注或表注写明。
从结果到机制的写法
每个实验小节按三句式收尾：现象 （曲线/表格显示什么）→ 机制 （模型里哪股力量导致）→ 含义 （对决策者意味着什么）。只有第一句的段落是图注，不是分析。
自检清单
Track record Benchmark 20 / 22
test cases passed, vs 7 / 22 without
Adoption
Users 0
Times used 0 View activity → Safety and source details
SkillSafety ✓ 3 of 3 passed ✓ Code scan 15 of 15 checks clean
0 critical · 0 warning · 0 info — scanner v4 — 8/10/2026
View 15 scanner checks live_secret remote_code_execution reverse_shell data_exfiltration ssrf_metadata agent_snooping anti_refusal destructive_commands obfuscated_unicode unicode_homoglyph opaque_blob suspicious_url tool_overreach trigger_abuse prompt_injection_phrasing
✓ Intent Safe
LLM judge · does the skill do what it claims
✓ Content Clean
LLM judge · policy & harmful-content review
Source Imported brycewang-stanford / awesome-journal-skills
Chinese-Journal-of-Management-Science-Skills/skills/cjms-numerical-experiments/SKILL.md
View on GitHub ↗ License MITImported 8/10/2026Version v1