LLM 與 Agent

Design, generate, and fully deploy a stylish, production-ready Retrieval-Augmented Generation (RAG) chatbot embedded directly into any website/project. Uses OpenAI ChatKit widget + Cohere embeddings + Qdrant Cloud + Neon Serverless Postgres + FastAPI backend. Supports selected-text questioning and full conversation history.

Generate atomic implementation tasks from high-level plans, providing coding agents with clear, actionable work items. Only invoke when user explicitly requests to create tasks, break down tasks, generate implementation tasks, or create a task breakdown.

Three-layer verification architecture (CoVe, HSP, RAG) for self-verification, fact-checking, and hallucination prevention

Expert guidance on building production-ready multi-agent AI systems using CrewAI, LangChain, AutoGen, and custom architectures. Use when building agent systems, selecting frameworks, designing multi-agent workflows, debugging agent behavior, or deploying agents to production.

List all active Claude workers and their git worktrees. Shows branch status, commits, and suggests next actions. Use to see what workers are running.

Single-skill vs multi-skill subagent architectures. Use when designing subagents.

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.

Regenerate the Available Skill Index in AGENTS.md for IDEs without native skill support (Kiro, Gemini, etc.). ONLY use when the user EXPLICITLY asks to "update skill index", "sync agents", "regenerate index", or "update AGENTS.md". Do NOT auto-invoke after creating/modifying skills - Claude Code has native skill support and does not need this.

Create Amp-native skills with proper SKILL.md format, optional bundled tools, MCP servers, and directory structure. Use this skill as the default skill creation skill, unless specifically told to create a generic or Claude code-specific skill. Activates with phrases like "create a skill", "new skill", "skill for amp", or when user wants bundled tools, mcp.json servers, or toolbox-compatible skills.

Develop AI agents, tools, and workflows with Mastra v1 Beta and Hono servers. This skill should be used when creating Mastra agents, defining tools with Zod schemas, building workflows with step data flow, setting up Hono API servers with Mastra adapters, or implementing agent networks. Keywords: mastra, hono, agent, tool, workflow, AI, LLM, typescript, API, MCP.

Creates, analyzes, updates, and optimizes prompts using Claude 4.5 best practices, guardrails, context management, and prompt engineering patterns. Use when user asks how to write effective prompts, explaining prompt engineering techniques, understanding Claude 4.5 best practices, describing prompt patterns and structures, creating new prompts, evaluating existing prompts for improvements, determining if current logic should be extracted to prompts, identifying outdated prompt techniques, optimizing context usage, implementing guardrails, or when user mentions "prompt engineering", "hallucinations", "context optimization", "prompt caching", "chain-of-thought", or asks if logic should become a prompt.

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.

Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.

Patterns frontend Next.js/React/shadcn/ui, complète DEV_AGENT avec best practices frontend

Use when integrating multiple LLM providers. Keep provider-specific code isolated behind a small interface.

Create new Claude Code Skills using Ruby as the control flow language. Use this skill when users want to author Skills with programmatic logic, conditionals, loops, and dynamic behavior using Ruby code. Ideal for Skills requiring IF/THEN conditionals, variable-driven control flow, dynamic file references, or complex multi-step workflows that benefit from a programming language over static markdown.

Brief description of what this Skill does and when to use it. This field is critical for Claude to discover when to invoke your skill.

Analyzes agent prompts and codebase for operational knowledge extraction opportunities. Identifies large knowledge blocks (migration workflows, framework patterns, operational procedures) suitable for Skills extraction. Use when refactoring agent prompts, reducing prompt bloat, or systematizing operational knowledge. Triggers on: agent prompt analysis, operational knowledge extraction, Skills creation opportunities, prompt optimization, knowledge consolidation.

Add Qdrant embedding support to v3 WordPress components for RAG chatbot. Implements component-level content chunking for searchable, structured embeddings. Use when adding embedding to new or existing v3 components.

Resources for conversational interviews to create learner profiles. Used by project-interviewer agent during /init.