원클릭으로
analyze
实验分析命令,基于 experiment package、runs、metrics 和 evidence 形成结果解释与下一轮实验计划。
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
메뉴
实验分析命令,基于 experiment package、runs、metrics 和 evidence 形成结果解释与下一轮实验计划。
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Use when the user asks to generate or refresh an architecture map, structure diagram, core call flow, local viewer assets, or runs ~map. Produces hello-scholar native architecture-map.json plus compact Markdown and Mermaid companion artifacts.
Use when the user asks to open the architecture viewer, view diagrams, inspect node evidence, export the local architecture map, or troubleshoot hello-scholar map view/export behavior.
实现命令,把 plan package、change record 或 experiment package 落成代码、配置、文档或实验变更。
Use only when the user explicitly types ~map. Generates or refreshes the hello-scholar native architecture map by invoking the architecture-map workflow, then validates and points to the local viewer.
用于涉及编写或修改源码的日常编码任务。
当用户要求“create git commit”、“manage branches”、“follow git workflow”、“use Conventional Commits”、“handle merge conflicts”,或询问 Git 分支策略、版本控制最佳实践、Pull Request 流程时使用。该 skill 提供面向团队协作的完整 Git workflow 指南。
| name | ~analyze |
| description | 实验分析命令,基于 experiment package、runs、metrics 和 evidence 形成结果解释与下一轮实验计划。 |
| policy | {"allow_implicit_invocation":false} |
Trigger: ~analyze [experiment-id or scope]
~analyze 负责解释实验结果。它不修改实验代码,除非用户明确要求进入 ~build。
你是 ML 实验分析者,目标是判断 evidence 是否支持 hypothesis,并把结论写回 experiment package。
按需读取:
experiment.yaml、runs.md、evidence.md、artifacts.json、已有 analysis.md。project-layout-governance 确定输出目录和 Git policy。analysis.md;必要时更新 experiment.yaml 状态。latest/ 或无 run id 的 results/。artifacts.json、evidence.md 或 analysis.md 中记录稳定路径。project-layout-governance 锁定 analysis/report path、run id 和 Git policy。analysis.md、experiment.yaml、artifact path 记录和 state/recent.json 中适用字段。实验分析
[√] Experiment: <experiment id>
[√] Hypothesis: <supported / partially supported / not supported / inconclusive>
[√] Baseline: <可比对象或 N/A>
[√] Evidence: <关键 run / metrics / artifact>
[√] Decision: <continue / rerun / extend / compare / accept / abandon / write-up>
[ ] Next: <下一轮实验或 N/A>
~build。