メニュー
Create or update an agent skill. Use when defining triggers, workflows, references, scripts, or assets for a reusable skill.
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/nateberkopec/dotfiles --skill skill-creator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。
Set up or audit Nate's standard dev environment in a project directory. Use for mise tasks, hk hooks, CI hardening, or standardized project tooling.
Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.
Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.
SOC 職業分類に基づく
SKILL.md を表示中
| name | skill-creator |
| description | Create or update an agent skill. Use when defining triggers, workflows, references, scripts, or assets for a reusable skill. |
| license | Complete terms in LICENSE.txt |
Establish concrete requests that should invoke the skill and what successful behavior looks like. For an existing skill, inspect observed failures or wasted work before editing. Include only knowledge or constraints that change the model’s default behavior.
Identify universal steps, branch-specific reference knowledge, deterministic operations, and output assets. When deciding package layout, read references/anatomy.md. Keep each rule in one source of truth and give every reference a conditional pointer.
For a new skill, run:
python ~/.claude/skills/skill-creator/scripts/init_skill.py <skill-name> --path <parent-directory>
Replace every placeholder and delete unused example resources. For an update, edit the existing package directly instead.
Write SKILL.md as concise verb-first instructions. Put invocation triggers in the description. Keep mandatory steps, safety constraints, branch selection, and completion criteria inline; disclose optional detail through references. Convert fixed pipelines into tested scripts.
python ~/.claude/skills/skill-creator/scripts/quick_validate.py <skill-directory>
python ~/.claude/skills/skill-creator/scripts/package_skill.py <skill-directory>
Fix every validation error. Exercise deterministic scripts with representative valid and invalid inputs. Use the skill on a real request, compare its behavior with the promised completion criteria, and prune no-op or duplicated instructions before considering it complete.