Master SQL query optimization, indexing strategies, and EXPLAIN analysis to dramatically improve database performance and eliminate slow queries. Use when debugging slow queries, designing database schemas, or optimizing application performance.
Write documentation following Metabase's conversational, clear, and user-focused style. Use when creating or editing documentation files (markdown, MDX, etc.).
Manages watcher processes that monitor Gmail, WhatsApp, filesystem, and other external sources. Use when starting, stopping, or monitoring watcher scripts, configuring process management, or troubleshooting watcher issues.
DDDアーキテクチャでHonoのバックエンドを新規実装または既存コードベースをリファクタリングする際のガイド。レイヤー構成（domain/application/infrastructure/controller）、依存方向、責務境界、Hono公式ミドルウェア/プラグインの活用、Hono OpenAPIプラグイン（hono-openapi）とzod-openapiの必須採用、Drizzle ORMの利用、トランザクション境界、エラーハンドリング、テスト指針、ESLint/Prettierの必須導入を含む。
作業指示ごとに .steering/ 配下へ requirements/design/tasklist を作成し、tasklist.md を唯一の進捗ソースとして実装を推進する。計画（作成）/実装（ループ）/振り返り（記録）の3モードで運用。「作業を計画して」「タスクリストを作成」「steering開始」「実装を進めて」などの依頼時に使用。
Use when writing or refactoring Python code for LangGraph v1.x + LangChain v1: Graph API + Functional API import paths, checkpoint/store usage, and migrations off deprecated prebuilts (create_react_agent -> langchain.agents.create_agent).
Create a safe checkpoint git commit on the current branch (auto-message; include agent log + safe untracked files; avoid secrets and bulky artifacts; track runs/**/{summary,report}.md only).
This skill ensures Claude always works in the JusticeHub codebase located at `/Users/benknight/Code/JusticeHub` and never accidentally works in the wrong repository.
エラーや不具合の調査と日本語レポート作成（軽いヒアリング付き・ファイル自動生成）
Use when working in Python with scikit-hep vector and Awkward Array to build vector records, register behaviors, compute deltaR or invariant masses, combine or boost vectors, or access vector properties from ak.zip records of type (Momentum3D/Momentum4D/Vector4D). Vector is only useful if you are also using Awkward.
Use when documenting Python research code usage, setup, and experiment reproduction.
Convert Notion pages or drafts into Hugo content for this repo, including front matter, shortcodes, glossary terms, and asset placement, and publish/preview with the Hugo CLI. Use when asked to import from Notion, format content to match FORMATTING.md, or create/update posts in content/.
Build evaluation frameworks for agent systems. Use when testing agent performance, validating context engineering choices, or measuring improvements over time.
Use when checking for data leaks, PII handling, and license risks in Python research code.
Data structure for annotated matrices in single-cell analysis. Use when working with .h5ad files or integrating with the scverse ecosystem. This is the data format skill—for analysis workflows use scanpy; for probabilistic models use scvi-tools; for population-scale queries use cellxgene-census.
[62] VALIDATE. Final self-check before delivery. Verify goal alignment, completeness, correctness, and identify residual risks. Produces quality score (0-100) and delivery status. Use when completing any significant work, before handoff, or when you need confidence that work is ready.
Next.js App Router における Server Actions と Route Handlers の使い分け、セキュリティ、および実装規約。
Ask the user to pick a response style and persist the preference.
This skill should be used when the user asks to "apply a layout", "arrange nodes", "organize the graph", "apply hierarchical layout", "use organic layout", "use tree layout", or mentions layout algorithms like "HierarchicalLayout", "OrganicLayout", "TreeLayout", "CircularLayout", "OrthogonalLayout", or "RadialLayout".
This skill should be used when the user asks to "build a graph from data", "load data into a graph", "import JSON into yFiles", "create nodes from an array", "use GraphBuilder", or mentions "data to graph", "import data", "business data", "GraphBuilder", "TreeBuilder", or "AdjacencyGraphBuilder".