メニュー
Resolve open decisions in the ledger by interviewing the user, one evidence-priced question at a time. Use when FOG.md has open KU decision items, or the user asks to be grilled about a plan.
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/leoncuhk/defog --skill fog-ask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。
During implementation, log deviations and newly discovered fog to the ledger instead of stopping or silently improvising. Use whenever implementing against a plan.
Generate a merge-gate quiz from the ledger, cross-checked against the diff — questions target what was actually uncertain.
Maintain a FOG.md unknowns ledger for a task. Use when a task has unresolved decisions or unverified constraints, when the user mentions unknowns or the fog ledger, or when another fog-* skill needs the ledger rules.
SOC 職業分類に基づく
SKILL.md を表示中
| name | fog-ask |
| description | Resolve open decisions in the ledger by interviewing the user, one evidence-priced question at a time. Use when FOG.md has open KU decision items, or the user asks to be grilled about a plan. |
Decisions belong to the user; facts belong to the territory. Before asking anything, re-triage every open item as fact or decision, and strike the facts via → lookup yourself.
For what remains, ask one question per turn, highest blast radius first — answers that would change the architecture before answers that pick a default. Every question must pass three gates: material (the answer changes the plan), grounded (points at evidence, not preference fishing), answerable (options plus a recommended default the user can wave through).
Format:
Question: … Why it matters: what changes between answers Evidence: file / doc / test citation Recommended: default + why. Silence = this.
Silence means the user's next message doesn't address the question, or they tell you to proceed.
Exits — both land the same way: convert every remaining item to [A], present the batch for veto.
If the user asks for everything at once, present the whole remaining frontier as one numbered round — recommendations attached — and treat unanswered items as silence.
Write each real answer to Resolved with its why. A defaulted answer goes to Assumed as [A], never to Resolved. Done when no open KU decisions remain.
