メニュー
Manage spaced repetition review for Obsidian knowledge cards. Use this skill when the user asks what to review today, wants to add cards to the review queue, or gives review feedback such as again/hard/good/easy for a card.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Create atomic Obsidian knowledge cards from user-provided ideas, sources, notes, or confirmed learning outcomes. Use when the user asks to save, record, create, or turn material into a reusable knowledge card, even without naming a card type.
对观点、论证与证据进行建设性审查。Use when the user asks to critique, validate, red-team, or stress-test a claim, card, draft, course, or project judgment for assumptions, evidence gaps, reasoning errors, counterexamples, boundaries, and next validation steps.
Coach active reading with ACTOR for books, articles, papers, transcripts, courses, and concepts. Use when the user wants a purposeful reading plan, comprehension or teach-back practice, a complete or single-stage ACTOR session, or to turn reading into tested understanding and action.
Process one Inbox clipping into useful knowledge, action, or an explicit rejection.
Welcome a new user to the DailyUp Second Brain Starter, give a concise guided tour, and safely remove template example files after explicit confirmation. Use only when the user explicitly invokes `/onboard`.
随机抽取知识库中的主题内容进行深度思考。当用户说"随机想法"、"随机思考"、"给我点灵感"、"来个有趣的话题"、"random thinking"或想要探索知识库中的惊喜内容时触发。技能会从知识卡片和知识文件中跨类型、跨主题地随机抽取内容，寻找隐含的关联和洞见。
| name | spaced-review |
| description | Manage active-recall spaced review for Obsidian knowledge cards. Use when the user asks what to review today, wants to enroll a card, or gives an explicit again/hard/good/easy rating for an identified review card. |
Use 主动提取: the user recalls before opening the card, receives correction, then gives a rating that schedules the next review.
Run:
python .agents/skills/spaced-review/scripts/spaced_review.py due --limit 20
Present the ordered cards using references/output-format.md. Let the user
select one card or explicitly end the session. When no cards are due, suggest
enrolling 3–5 core cards rather than bulk-enrolling the vault.
Completion: The user sees the due list and selects one card or ends the session.
For the selected card:
again, hard, good, or easy.Use these mappings when the user gives Chinese feedback: 忘了 → again; 想起一
部分/有点难 → hard; 能讲清楚 → good; 很熟且能迁移应用 → easy. If the
target card is not explicit in the current review turn, ask which card to
update.
Run:
python .agents/skills/spaced-review/scripts/spaced_review.py review "04_Knowledge/00_Cards/card_name.md" --rating good
Add --note only for user-provided feedback worth retaining. The script updates
frontmatter and appends the note to the card's SRS review log.
Completion: One identified enrolled card has an explicit rating, and the stored next due date and interval are reported.
Confirm the exact card path and choose core, normal, or low priority.
The default priority is normal; use core only for concepts the user needs to
retain long term.
Run:
python .agents/skills/spaced-review/scripts/spaced_review.py enroll "04_Knowledge/00_Cards/card_name.md" --priority core
Enrollment sets the card to learning, with a one-day interval and a next due
date tomorrow. Review requires prior enrollment.
Completion: The exact card path, priority, and next due date are confirmed from script output.
04_Knowledge/00_Cards/.references/review-fields.md when explaining stored review state.