메뉴
Analyze experiment results pulled by autodl-remote-pull-manifest (remote_artifacts) and generate a Hypothesis→Evidence→Next-Experiment (H→E→N) report. Use when you want evidence-first analysis from logs/metrics/vis and an actionable next-step plan to improve <主指标>.
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
Generate a minimal remote artifact pull manifest (results.json + train/logs + eval logs/metrics/vis) and place files under research_workspace/experiments/<ExpID>/remote_artifacts.
Generate a copy-paste remote (AutoDL/tmux) run snippet (git fetch/checkout/pull + scripts/train_eval_shutdown_autodl.sh). Use when you need a remote train→eval command block and must manually choose train/eval scripts from scripts/.
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
Ingest pulled remote artifacts under research_workspace/experiments/<ExpID>/remote_artifacts, summarize latest eval (summary.json), and update research_workspace/00-实验记录.md (index row + detailed section). Use when you want to record <主指标>/<辅助指标> into 00-实验记录.md after pulling remote artifacts.
Use when the input is a user-approved design doc or implementation plan and the agent needs to turn it into an executable issues CSV before handing off to CSV execution.
Use when executing an existing task CSV and the agent needs to push all actionable rows to closed-loop completion without stopping between issues.
| name | exp-analysis-hen |
| description | Analyze experiment results pulled by autodl-remote-pull-manifest (remote_artifacts) and generate a Hypothesis→Evidence→Next-Experiment (H→E→N) report. Use when you want evidence-first analysis from logs/metrics/vis and an actionable next-step plan to improve <主指标>. |
| metadata | {"short-description":"H→E→N experiment analysis from remote_artifacts"} |
Lightweight Claude mirror of the Codex skill. Full implementation and current script live under:
.codex/skills/exp-analysis-hen/.
research_workspace/experiments/<ExpID>/remote_artifacts/.research_workspace/experiments/<ExpID>/analysis.md.research_workspace/experiments/<ExpID>/analysis/.research_workspace/experiments/_cross_experiment/<RouteID>/.iter* or checkpoint*, report them by checkpoint and do not summarize them as mean±std or count them as seed evidence.mean±std is allowed and must be labeled as seed-level evidence.python3 .codex/skills/exp-analysis-hen/analyze.py --expid "E2025xxxx-xx"
python3 .codex/skills/exp-analysis-hen/analyze.py --expid "E2025xxxx-xx" --baseline-expid "E2025yyyy-yy"
Output: research_workspace/experiments/<ExpID>/analysis.md
docs/SKILLs/agent-skills-standard.md.codex/skills/exp-analysis-hen/SKILL.md (full documentation)