跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
数据与分析

materials-statistics

Use when auditing, revising, or drafting statistical reporting for materials-science manuscripts: sample-size and replicate definitions (specimens vs repeated measurements), ANOVA for factorial and Taguchi designs, response-surface regression, multiple-comparison corrections, effect sizes and uncertainty, error-bar and figure-statistics alignment, and reviewer comments about statistics. Trigger for statistics review, statistical analysis section, p-values, replicate counts, 统计审查、统计分析、 统计方法、方差分析、多重比较、样本量、重复数、置信区间、效应量、图注统计、 审稿人统计意见. Do not use for full raw-data reanalysis unless the user supplies data and explicitly asks for computation, for figure rendering, or for experiment planning without a reporting question.

cooleava1-gif · 已有静态检查记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add cooleava1-gif/Materials-Science-Skills --skill materials-statistics

技能说明


name: materials-statistics version: 1.0.0 stability: beta description: >- Use when auditing, revising, or drafting statistical reporting for materials-science manuscripts: sample-size and replicate definitions (specimens vs repeated measurements), ANOVA for factorial and Taguchi designs, response-surface regression, multiple-comparison corrections, effect sizes and uncertainty, error-bar and figure-statistics alignment, and reviewer comments about statistics. Trigger for statistics review, statistical analysis section, p-values, replicate counts, 统计审查、统计分析、 统计方法、方差分析、多重比较、样本量、重复数、置信区间、效应量、图注统计、 审稿人统计意见. Do not use for full raw-data reanalysis unless the user supplies data and explicitly asks for computation, for figure rendering, or for experiment planning without a reporting question.

Materials Statistics Reporting Router

Read manifest.yaml and its always_load files. Apply explicit direction > saved .materials/profile.yaml > neutral fallback, then resolve the task and study-design axes before loading any reference.

Blocking gates

  • replication-gate — if the independent experimental unit (n) cannot be determined from the supplied text, mark it AUTHOR_INPUT_NEEDED; never infer n from cell, image, spectrum, or reading counts.
  • invention-gate — never invent or silently repair p-values, degrees of freedom, confidence intervals, software versions, or correction methods.
  • boundary-gate — reporting and wording skill only; computation happens only on user-supplied data with an explicit request.

Routing protocol

  1. Classify the task (audit / rewrite / draft / reviewer-response support / figure-statistics alignment / data-backed check).
  2. Extract the design: groups, factors, levels, blocks, repeats, standards invoked (ASTM / ISO / GB), and the claimed inference.
  3. Resolve study_design from the manifest axis and load its fragment.
  4. Load references/statistical-reporting.md before drafting any Statistical analysis subsection.
  5. Check references/common-failure-modes.md when nested measurements, many comparisons, small samples, or curve regressions appear.
  6. Align figure statistics with references/figure-statistics.md when captions, error bars, or significance markers are involved.
  7. Run references/reviewer-checklist.md before final delivery.

Hand off experiment planning to materials-doe, dataset packaging to materials-data, plotting to materials-figure, and wording control to materials-polishing. This skill owns only the statistical reporting layer.