メニュー
Break a feature into small executable tasks with exact code and commands. Use when a spec or feature is ready and you need concrete implementation steps before coding — produces the task list arc-executing-tasks runs.
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/GregoryHo/arcforge --skill arc-writing-tasks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。
Use this skill whenever the user (an arcforge contributor) says they want to bump arcforge's version, cut a release, "ship vX.Y.Z", "準備發版", "ready to release", or any equivalent intent on the arcforge repo itself — even if they don't use the word "release". Runs the canonical release workflow: pre-flight checks → vault ingest → outdated-doc audit → CHANGELOG → 9-file version bump (incl. website) → commit/push/PR → post-merge tag. Contributor-only; do NOT trigger inside projects that merely install arcforge as a plugin.
Execute a task list with one fresh subagent per task plus a single task-reviewer each. Use when tasks need isolated contexts to avoid cross-task pollution; for in-session human-checkpoint execution use arc-executing-tasks.
Read-only advisory audit of an SDD spec family — design, spec, dag, and decision anchors.
SKILL.md を表示中
SOC 職業分類に基づく
| name | arc-writing-tasks |
| description | Break a feature into small executable tasks with exact code and commands. Use when a spec or feature is ready and you need concrete implementation steps before coding — produces the task list arc-executing-tasks runs. |
| category | sdd |
| status | promoted |
| argument-hint | <feature-name> |
Break features into bite-sized tasks (2-5 minutes each) with exact code and commands. Each task is independently executable.
docs/tasks/<name>-tasks.md| Too Vague | Just Right |
|---|---|
| "Set up auth" | "Create User dataclass in src/types.py" |
| "Add tests" | "Write test_login_invalid_password in tests/auth/" |
| "Implement login" | "Add password hash check in login()" |
Output to docs/tasks/<feature-name>-tasks.md. Each task is TDD-ordered: write failing test → run (expect FAIL) → implement → run (expect PASS) → commit.
# <Feature Name> Tasks
> **Goal:** [One sentence describing what this delivers]
> **Architecture:** [2-3 sentences about approach]
> **Tech Stack:** [Key technologies/libraries]
> **For Claude:** Use arc-agent-driven or arc-executing-tasks to implement.
## Context
[Scene-setting from feature spec]
## Tasks
### Task 1: [Name]
**Files:**
- Create: `exact/path/to/file.py`
- Test: `tests/exact/path/to/test.py`
**Step 1: Write failing test**
\`\`\`python
def test_specific_behavior():
...
\`\`\`
**Step 2: Run test**
Run: `pytest tests/path/test.py::test_name -v`
Expected: FAIL
**Step 3: Implement**
\`\`\`python
def function():
...
\`\`\`
**Step 4: Verify**
Run: `pytest tests/path/test.py::test_name -v`
Expected: PASS
**Step 5: Commit**
`git commit -m "feat: add specific feature"`
### Task 2: ...
✅ Tasks written → docs/tasks/-tasks.md
⚠️ Writing tasks blocked
Vague = break down further.
Hand off to one of:
arc-agent-driven — automated execution, fresh subagent per task. Best for walk-away batch runs where each task is well-scoped.arc-executing-tasks — human-in-the-loop mode with checkpoint prompts between tasks. Best when the task list carries judgment calls or when you want to review each step.Both read docs/tasks/<name>-tasks.md as input. Default to arc-agent-driven.