Menu
checklist step of mspec workflow — produce checklist.md via subagent
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
archive step of mspec workflow — deterministic merge into SoT spec and archive move
archive ステップ完了後に Lessons を分析し constitution.md への追加提案を生成するサブエージェント
archive ステップ完了後に Next Steps を評価し優先度付きのチェンジ登録提案を生成するサブエージェント
proposal step of mspec workflow — clarify intent and write proposal.md
archive step of mspec workflow — deterministic merge into SoT spec and archive move
delta step of mspec workflow — generate Delta Spec with auto-numbered FR-NNN
| name | mspec-checklist |
| description | checklist step of mspec workflow — produce checklist.md via subagent |
| when_to_use | User runs /mspec:checklist, or workflow auto-continues to checklist |
mspec status --change <change-dir> --json first.mspec continue --change <change-dir> --json to fetch subagent_prompt.requires (specs/*/spec.md, design.md).mspec-checklist-auditor subagent (Task tool) — it should also scan related capability SoT specs for regression risk.checklist.md (Delta Spec Coverage / Source-of-Truth Regression / Constitution).
5a. readme.md の ## Artifacts 節の - [ ] checklist.md を - [x] checklist.md に更新する。mspec validate --change <change-dir>. validate が失敗した場合は - [x] checklist.md を - [ ] checklist.md にロールバックする。block: false — auto-continue.After the subagent completes (step 4), record the run:
mspec agent-run record checklist \
--change <change-name> \
--bytes <sum-of-input-artifact-bytes> \
--artifacts <space-separated-artifact-paths>
Example: mspec agent-run record checklist --change 2026-05-25-my-feature --bytes 12340 --artifacts proposal.md specs/my-cap/spec.md design.md
This appends one JSONL entry to changes/<change>/.agent-runs.jsonl. Do NOT include prompt text or file contents.
mspec validate --change <change> — アーティファクト整合性チェックmspec anchor check --change <change> — アンカー解決確認mspec verify llm --change <change> — FR-IDごとのLLM評価プロンプトを生成してClaude Codeで実行（C2強化）verify: fr-NNN アノテーションがDelta Specの全FR-IDを網羅しているか手動確認このスキルの実行で発生した学習候補を記録する:
<!-- LEARNING: <パターン説明> | source: <FR-ID> | confidence: low|medium|high -->
mspec learn コマンドが archive 済み changes からこれらを収集してpost-condition候補をproposeする。