メニュー
Reveal the solution for the active Mentor Mode objective after the learner explicitly asks. Use only when the user runs /reveal or explicitly asks to be shown the solution code.
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
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。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/bradtraversy/mentor-mode --skill revealコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。
Enable Mentor Mode on this repository. Use when the user runs /mentor-init, asks to set up mentor mode, enable learning mode, or initialize the guided learning workflow on a project.
End a Mentor Mode session with the explain-back gate, ledger and curriculum updates, a session log, and a next-session preview. Use when the user runs /mentor-wrap, says wrap up, end session, or done for today in a Mentor Mode repo.
Start a Mentor Mode learning session. Use when the user runs /mentor, says start a session, begin today's lesson, or let's learn in a Mentor Mode repo.
SOC 職業分類に基づく
SKILL.md を表示中
| name | reveal |
| description | Reveal the solution for the active Mentor Mode objective after the learner explicitly asks. Use only when the user runs /reveal or explicitly asks to be shown the solution code. |
Reveals are for genuine walls, not friction. Confirm once, briefly: this goes on the re-derive queue and comes back in a future warmup. If the learner has not yet made a real attempt and walked the hint ladder, offer the next hint level first - but an explicit "reveal it" wins. Never nag twice.
mentor/guard.json: set reveal.active true, reveal.objectiveId to the active objective id, grantedAt to the current ISO timestamp.Write or show the solution for the specific thing the learner is stuck on - the objective's scope, nothing extra. Explain it line by line at the learner's level: what each part does and why this shape rather than the obvious alternative. If writing into protected files, keep it minimal and idiomatic.
Immediately after delivering the solution: set reveal.active false and reveal.objectiveId null in mentor/guard.json. Reveals are one-shot. As a backstop, the hook ignores any reveal whose grantedAt is older than 2 hours.
The objective stays in-progress until the learner can explain the revealed code back. Do that explain-back now. The code returns in a future warmup for re-derivation from memory.