YasuakiOmokawa
UserAgent skills for improve development
Categories
Indexed Skills (16)
qa-ui
実装完了後の UI を検証したいとき、または「UI を確認して」「QA して」「画面の動作確認」と頼まれたときに使用。ChromeDevTools MCP を使い、独立コンテキストの QA エージェントが画面を操作して AC 単位で pass/fail を判定し、Major/Minor 不合格は最小修正して再検証するループを回す (最大3ラウンド、Critical は即エスカレート)。
create-jira-issues
ユーザーストーリーマップやタスク分解プランからJiraチケットを一括作成する場合に使用。ストーリーマップ完成後のチケット化、プランファイルからの一括作成、計画済みストーリー・タスクのJiraへの移行が必要な場合にトリガーされる。
create-pr
Creates a Conventional-Commits draft PR from the current branch with generated title, body, labels, and milestone, without confirmation. Use when the user says "PR を作って" / "draft PR" / "PR 作成して", optionally with `[base-branch]` argument.
define-acceptance-criteria
Fills a matrix of 3 required categories (normal, error, edge) by controlled-vocabulary perspectives to enumerate acceptance criteria into the analysis file. Use when in plan mode before /mece-plan-review, when the user asks to write AC for a plan, or when an AC matrix is needed as MECE input.
dry-ssot-text
Collapses repeated explanations of one concept into a single source of truth plus cross-references, while keeping navigation aids like TOCs and progress tables. Use when an AI-generated document (plan / design doc / RFC / PR description) has grown long with the same concept explained in multiple places.
finalize-plan
Turns AC and MECE results from the analysis file into a branch strategy, PR split, and manual/auto QA steps appended to the plan file. Use when the user has completed `/define-acceptance-criteria` + `/mece-plan-review` and is about to move from plan mode into implementation.
map-user-stories
設計書・プロジェクト仕様・Jira epic等からユーザーストーリーマップを作成し、タスク分解・スプリント計画まで行う場合に使用。新しいプロジェクトフェーズの計画、設計書の分析、大きな機能の実装単位への分解が必要な場合にトリガーされる。
mece-plan-review
Verifies acceptance criteria for MECE coverage with parallel spec/code/wiki analysts and a fresh red-team judge, recording coverage gaps and duplicates in the analysis file. Use when AC is already defined in the analysis file via /define-acceptance-criteria and MECE verification is required before implementation.
polish-before-commit
Auto-fixes convention and pattern-consistency issues, runs lint, and aggregates remaining judgment calls before stopping for the user. Requires the Claude `feature-dev` plugin for the final review step and halts with install guidance when it is missing. Use when finalizing a branch, just before `git commit` or `/create-pr`, or whenever the user says "仕上げて" / "polish" / "コミット前チェック".
purge-private-vocab
Detects local-plan coinages, abbreviations, and number labels in reader-facing text and rewrites them so readers without the source plan can follow. Use after generating PR description, Jira ticket, design doc, RFC, or other reader-facing text from a local plan/spec file, when readers don't share the source plan.
review-code-quality
Analyzes a diff across cohesion, coupling, and readability (plus business-impact for domain-attribute changes), auto-applies only mechanically safe readability-axis critical/major fixes (cohesion/coupling/business-impact findings are always handed off), and routes design-judgment items to /polish-before-commit. Use when finishing self-review of an implementation, before requesting PR review, or when a diff updates a domain model attribute (`plan_code` / `role` / `status` 等).
review-design
Reviews code placement and pattern choices before implementation using selected architecture reviewers (anti-pattern plus DDD, Hexagonal, Clean) and a mandatory adversarial critique, editing the plan directly on fatal findings. Use when starting a new feature, adding a file or module, deciding "where should this code live", or when the user requests a design review with `/review-design`.
set-jira-story-points
JiraキーとStory Pointsのマップデータを受け取り、Jiraチケットに一括でStory Pointsを設定する。「ストーリーポイント設定」「SP設定」「story points」などのキーワードでトリガーされる。
translate-to-vision-story
プロジェクト活動 (commits/PRs/README/ADR) を `~/.claude/skills-config/vision.md` のビジョン要素と照合し、対話型 draft → revise loop で Zenn 記事下書きを生成する。プロジェクト単位の物語化・キャリアブランディング・月次記事執筆時に使用。
iterate-with-prototypes
Use when starting a complex feature where a PRD or spec exists but load-bearing assumptions (technical feasibility, UX effect, reuse of an existing API or data structure) are still unverified, most implementation is done by an AI agent, the work spans several design docs and a PR chain, AND the change is reversible with a small blast radius. Do NOT use when the core risk is a hard-to-reverse decision (DB schema / migration, public API contract, cross-team boundary) — use design-first there. Symptom — about to design a complex feature on paper before any of it is proven to run, or confident-but-wrong design propagating across documents.
model-data
要求文書からDBML形式のER図を生成し、SQLアンチパターンを検出。DB設計、ER図作成、スキーマ正規化、既存設計レビュー時に使用。
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.