SOC 직업 분류 기준
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.
npx skills add https://github.com/natolambert/rlhf-book --skill pre-submit-pr명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.
로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.
SKILL.md 표시 중
| name | pre-submit-pr |
| description | Validate changes before PR submission |
Validate changes before submitting a pull request.
/pre-submit-pr
Lint & format (for changes under code/):
cd "$(git rev-parse --show-toplevel)/code" && uvx ruff@0.14.5 check . && uvx ruff@0.14.5 format --check .
Auto-fix with uvx ruff@0.14.5 check --fix . and uvx ruff@0.14.5 format . if issues are found. The ruff@0.14.5 pin matches .github/workflows/lint.yml; unpinned uvx ruff can diverge from CI.
Changelog (for changes under code/):
CI requires code/CHANGELOG.md to be modified for any PR touching code/ (format is convention, not enforced). Check that an entry exists under ## Unreleased for this PR, format: - YYYY-MM-DD: [PR #N](https://github.com/natolambert/rlhf-book/pull/N) description.
Run code review (for significant changes):
Invoke pr-review-toolkit:review-pr for deeper analysis.
Include findings in the report under a "### Code Review" section.
Summarize PR readiness
## Pre-Submit PR Report
### Automated Checks
| Check | Status | Details |
|-------|--------|---------|
| Ruff lint | PASS/FAIL | [details] |
| Ruff format | PASS/FAIL | [details] |
| Changelog | PASS/MISSING | [details] |
### Code Review
[pr-review-toolkit:review-pr findings, if run]
### Verdict: READY FOR PR / ISSUES TO ADDRESS
### Summary for PR Description
[2-3 sentences summarizing changes]
These block PR submission:
code/CHANGELOG.md (CI-enforced: file must be modified)Note in PR but don't block:
Plan, run, and report a small RLHF Book code experiment.
Turn the first page of a paper (arXiv or any PDF) into a slide-ready PNG for a colloquium deck's 30-40% right column. Use when a slide cites a landmark paper and a screenshot of the paper itself is the best visual.
Extract per-question (MM:SS) timestamps from a recorded course video by OCR-ing the slide counter, then add YouTube deep links to the course page's Q&A question lists. Use after a Q&A or lecture recording is published, or when adding/refreshing "Show questions" timestamp links in book/templates/course.html.