jpskill.com
📦 その他 コミュニティ

spec-kit-tasks

Use when an approved Spec Kit `plan.md` + `spec.md` must be decomposed into dependency-ordered `tasks.md`, or when `tasks.md` is missing/stale after planning or reconciliation changes.

⚡ おすすめ: コマンド1行でインストール(60秒)

下記のコマンドをコピーしてターミナル(Mac/Linux)または PowerShell(Windows)に貼り付けてください。 ダウンロード → 解凍 → 配置まで全自動。

🍎 Mac / 🐧 Linux
mkdir -p ~/.claude/skills && cd ~/.claude/skills && curl -L -o spec-kit-tasks.zip https://jpskill.com/download/10653.zip && unzip -o spec-kit-tasks.zip && rm spec-kit-tasks.zip
🪟 Windows (PowerShell)
$d = "$env:USERPROFILE\.claude\skills"; ni -Force -ItemType Directory $d | Out-Null; iwr https://jpskill.com/download/10653.zip -OutFile "$d\spec-kit-tasks.zip"; Expand-Archive "$d\spec-kit-tasks.zip" -DestinationPath $d -Force; ri "$d\spec-kit-tasks.zip"

完了後、Claude Code を再起動 → 普通に「動画プロンプト作って」のように話しかけるだけで自動発動します。

💾 手動でダウンロードしたい(コマンドが難しい人向け)
  1. 1. 下の青いボタンを押して spec-kit-tasks.zip をダウンロード
  2. 2. ZIPファイルをダブルクリックで解凍 → spec-kit-tasks フォルダができる
  3. 3. そのフォルダを C:\Users\あなたの名前\.claude\skills\(Win)または ~/.claude/skills/(Mac)へ移動
  4. 4. Claude Code を再起動

⚠️ ダウンロード・利用は自己責任でお願いします。当サイトは内容・動作・安全性について責任を負いません。

🎯 このSkillでできること

下記の説明文を読むと、このSkillがあなたに何をしてくれるかが分かります。Claudeにこの分野の依頼をすると、自動で発動します。

📦 インストール方法 (3ステップ)

  1. 1. 上の「ダウンロード」ボタンを押して .skill ファイルを取得
  2. 2. ファイル名の拡張子を .skill から .zip に変えて展開(macは自動展開可)
  3. 3. 展開してできたフォルダを、ホームフォルダの .claude/skills/ に置く
    • · macOS / Linux: ~/.claude/skills/
    • · Windows: %USERPROFILE%\.claude\skills\

Claude Code を再起動すれば完了。「このSkillを使って…」と話しかけなくても、関連する依頼で自動的に呼び出されます。

詳しい使い方ガイドを見る →
最終更新
2026-05-18
取得日時
2026-05-18
同梱ファイル
1
📖 Claude が読む原文 SKILL.md(中身を展開)

この本文は AI(Claude)が読むための原文(英語または中国語)です。日本語訳は順次追加中。

Spec Kit Tasks

Generate an implementation-ready tasks.md from Spec Kit design artifacts.

When to Use

  • plan.md and spec.md exist and you need executable tasks for implementation.
  • tasks.md is missing, stale, or not aligned with current plan/spec artifacts (including post-spec-kit-reconcile updates).
  • You need user-story-scoped task phases with clear dependency and parallelization signals.

When Not to Use

  • The feature spec is missing (spec-kit-specify first).
  • High-impact ambiguity still blocks design decisions (spec-kit-clarify first).
  • Technical design artifacts are not finalized (spec-kit-plan first).
  • You are executing tasks rather than generating them (spec-kit-implement).

Router Fit

  • Primary route from spec-kit after spec-kit-plan.
  • Must complete before spec-kit-implement.
  • Supports spec-kit-analyze and spec-kit-reconcile by producing structured implementation intent.

Preconditions

  • Run from repository root (or a subdirectory inside it).
  • Active feature context resolves to one specs/<feature>/ directory.
  • plan.md exists and reflects the latest approved design.

Workflow

  1. Resolve feature paths and prerequisite gate:

    • Run scripts/check-prerequisites.sh --json exactly once.
    • Parse FEATURE_DIR and AVAILABLE_DOCS.
    • Derive:
      • FEATURE_SPEC = FEATURE_DIR/spec.md
      • IMPL_PLAN = FEATURE_DIR/plan.md
      • TASKS = FEATURE_DIR/tasks.md
    • If spec.md is missing, stop and route to spec-kit-specify then spec-kit-plan.
  2. Load generation inputs:

    • Required: plan.md, spec.md.
    • Optional (when present): research.md, data-model.md, contracts/, quickstart.md.
    • Template preference:
      • {REPO_ROOT}/templates/tasks-template.md
      • {REPO_ROOT}/.specify/templates/tasks-template.md
      • fallback: assets/tasks-template.md
  3. Extract planning context:

    • From plan.md: stack, architecture, constraints, project structure.
    • From spec.md: prioritized user stories, acceptance criteria, independent test intent.
    • From optional docs: shared entities, external interfaces, and setup decisions.
  4. Build task phases:

    • Phase 1: Setup.
    • Phase 2: Foundational prerequisites blocking all stories.
    • Phase 3+: one phase per user story in priority order.
    • Final phase: Polish/cross-cutting concerns.
  5. Generate tasks in strict checklist format:

    • Required pattern: - [ ] T### [P?] [US#?] Action with file path.
    • Include [US#] only for user-story phases.
    • Include [P] only when tasks can run safely in parallel.
    • Include exact file paths in every implementation/test task description.
    • Add test tasks only when explicitly requested by spec/user.
  6. Validate coverage and ordering before writing:

    • Every user story has independently testable tasks.
    • Dependencies are explicit and respect phase order.
    • Every task matches format requirements (checkbox, ID, labels, file path).
    • No orphaned entities/contracts without mapped tasks.
  7. Write tasks.md:

    • Preserve heading order from the selected template.
    • Replace template placeholders and sample content with concrete feature tasks.
  8. Report completion:

    • Absolute tasks.md path.
    • Total task count and per-story counts.
    • Parallel opportunities.
    • Suggested MVP slice (typically first priority story).
    • Readiness handoff for spec-kit-implement.

Task Format Guidance

Format Components

  1. Checkbox: ALWAYS start with - [ ] (markdown checkbox).
  2. Task ID: Sequential number (T001, T002, T003, ...) in execution order.
  3. [P] marker: Include ONLY if the task is parallelizable (different files, no dependencies on incomplete tasks).
  4. [Story] label: REQUIRED for user story phase tasks only.
    • Format: [US1], [US2], [US3], etc. (maps to user stories from spec.md)
    • Setup phase: no story label
    • Foundational phase: no story label
    • User story phases: MUST have story label
    • Polish phase: no story label
  5. Description: Clear action with exact file path.

Examples

  • CORRECT: - [ ] T001 Create project structure per implementation plan
  • CORRECT: - [ ] T005 [P] Implement authentication middleware in src/middleware/auth.py
  • CORRECT: - [ ] T012 [P] [US1] Create User model in src/models/user.py
  • CORRECT: - [ ] T014 [US1] Implement UserService in src/services/user_service.py
  • WRONG: - [ ] Create User model (missing Task ID and Story label)
  • WRONG: T001 [US1] Create model (missing checkbox)
  • WRONG: - [ ] [US1] Create User model (missing Task ID)
  • WRONG: - [ ] T001 [US1] Create model (missing file path)

Output

  • tasks.md under the active feature directory with dependency-ordered, story-scoped tasks.
  • Completion summary with task counts, dependency highlights, and MVP recommendation.

Key Rules

  • Organize tasks by user story so each story can be implemented and validated independently.
  • Keep task granularity implementation-ready: no vague one-liners and no giant umbrella tasks.
  • Preserve execution truth in IDs and dependency order (T001, T002, ...).
  • Treat foundational work as blocking unless it is explicitly independent and parallel-safe.
  • Never invent tests by default; include test work only when requested.

Common Mistakes

  • Generating tasks from spec.md alone without using plan.md constraints.
  • Copying sample tasks from the template instead of replacing them.
  • Missing file paths or story labels in user-story phases.
  • Marking conflicting tasks as [P] even though they touch the same files/dependencies.
  • Producing tasks that cannot satisfy each story's independent test criteria.

References

  • references/spec-kit-workflow.dot for where task generation fits in the full Spec Kit sequence.
  • scripts/check-prerequisites.sh
  • assets/tasks-template.md
  • https://github.com/github/spec-kit/blob/9111699cd27879e3e6301651a03e502ecb6dd65d/templates/commands/tasks.md