菜单
Use when a run is finished and the report is written, after the report is written, to record one reusable lesson for the next run in this field. Covers what counts as a lesson worth passing on, what must never be passed on, and how to write it.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/tangxiangru/AutoR --skill record-what-you-learned命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Use when writing results and discussion, and when a task or a reviewer asks why an effect happens rather than whether it does. Covers the difference between reporting an effect and accounting for it, and what a mechanism claim needs behind it.
Use when the supplied data looks synthetic, degraded, incomplete or wrong, and whenever you are tempted to reframe the study around what you found about the inputs, the harness or the evaluation. Covers what to do with a real data problem without losing the study.
Use when planning figures, at study design and again before writing. Covers the figures a paper in this field is expected to contain, why an original figure does not substitute for a standard one, and how to decide what to draw first.
基于 SOC 职业分类
正在显示 SKILL.md
| name | record-what-you-learned |
| description | Use when a run is finished and the report is written, after the report is written, to record one reusable lesson for the next run in this field. Covers what counts as a lesson worth passing on, what must never be passed on, and how to write it. |
You hit things this run that were in no prompt: an archive that stores its axis in an unexpected order, a reference implementation needing an undocumented flag, a check that caught a mistake you would otherwise have shipped, a step the field treats as obvious and no instruction mentioned.
The next run in this field hits the same thing unless you write it down.
Run this once, at the end:
python3 -c "
import sys; sys.path.insert(0, '<AUTOR_ROOT>')
from src.skill_evolution import record_note
note, problems = record_note(
discipline='<the field, e.g. earth>',
title='<short, routable, what the lesson is about>',
body='''<what you hit, and what to do instead next time>''',
learned_in='<this task id>')
print(problems or 'recorded')
"
A good note is one paragraph answering: what surprised you, how it shows up, and what to do instead. Write it for someone competent who has not seen this corpus.
No results. Not your numbers, not the paper's, not "it came out around X". A note travels to a different task, and a finding that travels is contamination — it invites the next run to expect an answer instead of measuring one. The recorder refuses notes containing measured values, and that refusal is not an obstacle to work around.
Nothing you did not hit. A guess about what might help is prose, and prose accumulates until nobody reads the pool. A run that learned nothing transferable records nothing. That is a valid outcome.
Not five. The pool is capped, and a run filing five pushes out four another run earned. Pick the one you most wish you had known at the start.