LLM & Agents

Reviews Claude skills against Anthropic's official design philosophy and specifications. Use when asked to review, validate, or check if a skill follows best practices.

World-class prompt engineering skill for LLM optimization, prompt patterns, structured outputs, and AI product development. Expertise in Claude, GPT-4, prompt design patterns, few-shot learning, chain-of-thought, and AI evaluation. Includes RAG optimization, agent design, and LLM system architecture. Use when building AI products, optimizing LLM performance, designing agentic systems, or implementing advanced prompting techniques.

効果的なスキルを作成するためのガイド。ユーザーが専門知識、ワークフロー、ツール統合でClaudeの能力を拡張する新しいスキルを作成（または既存のスキルを更新）したい場合に使用します。

子エージェントの起動・指示・報告・安全運用を標準化するスキル。codex でのタスク分解、確認判断、チェックリスト、テンプレートを提供し、リーダー/実装/レビュー各エージェントが迷わず動けるようにする。

Generate production-ready Workscript workflow JSON files for the Workscript Agentic Workflow Engine with built-in defensive guards and complexity detection. Use when asked to create workflows, automations, data pipelines, or generate workflow JSON. IMPORTANT - For complex requests that would result in deeply nested or overly complex workflows, this skill will STOP and suggest developing new custom nodes first using /new-node, then return to create a simpler, more linear workflow. All generated workflows include data validation (validateData node) for structured JSON outputs, input guards, array length checks, and error handling edges. Fetches up-to-date node documentation from the Reflection API when available. Outputs validated .json files to the sandbox prompts folder. Also suitable for Claude Code subagents needing to compose workflows programmatically.

Coordinates agent assignments, validates merge states, and manages ticket lifecycle.

Customer support and issue resolution

Manage Cloudflare R2 object storage buckets and objects using Wrangler CLI. Use when working with R2 buckets, uploading/downloading files, configuring custom domains, CORS, lifecycle policies, or managing R2 storage. Trigger keywords: R2, bucket, object storage, Cloudflare storage, wrangler r2.

Agent Skill: Systematic TYPO3 extension upgrades to newer LTS versions. Covers Extension Scanner, Rector, Fractor, PHPStan, and testing. Use when upgrading extensions to newer TYPO3 versions or fixing compatibility issues. By Netresearch.

Use when creating, designing, or building Claude Code skills. Auto-activates for "create a skill", "build a skill", "skill for X", or "make this into a skill". Provides complete 6-step creation protocol, embedded Anthropic documentation, working examples, and templates. Never need to fetch external docs again.

Presentation creation, editing, and analysis. When Claude needs to work with presentations (.pptx files) for: (1) Creating new presentations, (2) Modifying or editing content, (3) Working with layouts, (4) Adding comments or speaker notes, or any other presentation tasks

Testing and quality validation for Flutter code.USE WHEN: creating tests, verifying coverage, running E2E tests,validating code quality, checking test coverage thresholds.Minimum coverage: 80% overall, 95% services, 70% widgets.Examples:<example>Context: The user implemented a new feature.user: "I've just implemented the new itinerary creation service"assistant: "I'll use testing-agent to generate comprehensive tests and ensure quality."<commentary>New code needs testing - use testing-agent.</commentary></example><example>Context: The user wants to run E2E tests.user: "Run the E2E tests to validate the login flow"assistant: "I'll use testing-agent to execute E2E tests with MCP Chrome DevTools."<commentary>E2E testing is a testing-agent responsibility.</commentary></example>

Provides creative thinking prompts using concrete metaphors, dramatic scenarios, and strategic frameworks. Use when the user needs inspiration, is stuck on a problem, or wants to validate ideas from first principles.

ai-tools同期 - リポジトリと~/.claude/間の設定ファイル同期。to-local/from-local/diffモード。

Build, test, and distribute Claude Code plugins with slash commands, agents, skills, hooks, MCP servers, and LSP servers. MUST be loaded when creating, reviewing, debugging, or distributing plugins. Use PROACTIVELY when user mentions plugins, extensions, marketplaces, or wants to add custom commands/agents to Claude Code.

Index of AI agent skills and how to use them when implementing features in this repo.

Use when working with Neon serverless PostgreSQL database. Covers database setup, CLI commands, Drizzle ORM integration, schema migrations, connection pooling, and user data storage patterns for authentication and chat history.

Improve prompts for AI agents and Telegram bots using OpenAI's prompt engineering best practices. Analyzes clarity, specificity, context, and output format. Returns structured improvements.

Use when executing decomposed plans with parallel batches - dispatches up to 2 fresh subagents per batch with code review between batches, enabling fast parallel iteration with quality gates

Guide for implementing features in ClaudeBar following architecture-first design, TDD, rich domain models, and Swift 6.2 patterns. Use this skill when:(1) Adding new functionality to the app(2) Creating domain models that follow user's mental model(3) Building SwiftUI views that consume domain models directly(4) User asks "how do I implement X" or "add feature Y"(5) Implementing any feature that spans Domain, Infrastructure, and App layers