LLM & Agents

驗證 maihouses 專案的 TypeScript/React 代碼品質，確保符合 CLAUDE.md 的最高規格標準。當修改或審查代碼時自動使用。

LLM-only skill that takes a textual description of a new skill and generates a complete minimal file plan for a new MOVA skill (schemas, envelope, manifest, prompt profile, runtime binding, docs, cases).

Generates new Claude Code custom slash command prompts following the established template structure. Use when the user asks to create a new slash command, generate a command prompt, or build a custom command for .claude/commands/

Expert guidance on unit testing for the Pomodoro Time Tracker. Activates when working with tests, test coverage, or testable code patterns.

Expert in Shelby Protocol TypeScript SDK for decentralized storage on Aptos. Helps with blob uploads/downloads, session management, micropayment channels, multipart uploads, and SDK integration for Node.js and browser environments. Triggers on keywords ShelbyNodeClient, ShelbyClient, @shelby-protocol/sdk, Shelby SDK, decentralized blob storage, Shelby upload, Shelby download, Shelby session.

JavaScript/TypeScript stack configuration - pnpm, prettier, eslint, vitest with 96% coverage threshold

Guide for using and supporting the AGENTS.md standard in VS Code. Use this when asked about AGENTS.md, custom instructions, or repo-level AI agent configuration.

Practice radical candor by delivering only verified, tested code with data-backed decisions, immediate problem flagging, and honest status communication. Use this skill when making claims about code behavior, reporting system status, identifying risks, documenting missing coverage, or challenging assumptions. Applies to all development activities requiring factual accuracy, evidence-based assertions, gap ownership, and closing the loop on delivered work through validation and independent review.

MongoDB database exploration and querying. Use when you need to understand database structure, view existing data, check collection schemas, count documents, or run queries to investigate the database state. (project)

Agent orchestration patterns for the Pomodoro Time Tracker. Activates for multi-agent tasks, test failures, or complex implementations.

Automatically detect when features change and remind to update the feedback assistant knowledge base. CRITICAL for keeping the AI assistant accurate and preventing user frustration from outdated information.

NanoBanana（Google Gemini画像生成）向けの高品質プロンプトを生成。画像生成、編集、フェイススワップ、背景変更、キャラクター一貫性のプロンプト作成を支援。

Terragrunt wrapper for Terraform with DRY configurations. Enforces plan-before-apply workflow. Use when working with terragrunt.hcl files.

Gate 6 - Verify tests exist and cover critical paths. Issues result in WARNINGS (encourages tests, doesn't block).

This skill should be used when creating or updating implementation documentation, task breakdowns, verification steps, or phase planning documents. It provides standards and templates for consistent, professional documentation throughout the MTG Agent project.

LLMs, prompt engineering, RAG systems, and AI application development

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

Agent-to-Agent (A2A) executor implementation patterns for task handling, execution management, and agent coordination. Use when building A2A executors, implementing task handlers, creating agent execution flows, or when user mentions A2A protocol, task execution, agent executors, task handlers, or agent coordination.

Manages Claude Skills lifecycle - creating, updating, maintaining, and versioning project-specific and cross-IDE knowledge bases. Invoke when user wants to create new skills, update existing ones, or sync skill content across tools.

expo-secure-store patterns for sensitive data. Use when storing tokens and credentials.