awesomemcp-builder
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
locomo-benchmark
Run LoCoMo benchmark for long-term conversational memory
hosted-mcp-server
Creates and configures hosted Model Context Protocol (MCP) server connections for OpenAI Agents SDK
session-management
Initialize work sessions with context loading, recent history review, and co-pilot agent spawning. Use when starting work sessions, switching between projects, or needing focused context for deep work.
data-layer-architecture
Bronze Layer（LLM抽出ログ層）とGold Layer（確定データ層）の2層アーキテクチャ設計。LLM抽出結果の履歴管理と人間修正の保護を実現。抽出処理の実装、ExtractionLogの使用、is_manually_verifiedフラグの扱いに関するガイダンスを提供。
skill-creator
创建有效技能的指南。当用户想要创建新技能（或更新现有技能）以通过专业知识、工作流程或工具集成来扩展 Claude 能力时，应使用此技能。
moai-cc-settings
Claude Code settings management, preference customization, and user experience optimization. Use when customizing Claude Code behavior, managing user preferences, or optimizing development experience.
ac-stop-hook-analyzer
Analyze context and decide on continuation via Stop hook. Use when determining if work should continue, analyzing completion status, making continuation decisions, or implementing the Two-Claude pattern.
rag-design-patterns
Automatically applies when building RAG (Retrieval Augmented Generation) systems. Ensures proper chunking strategies, vector database patterns, embedding management, reranking, and retrieval optimization.
exps-performance
Documentation for the main LLM performance experiments (exps_performance)
lcql-query-builder
Activate when the user needs help writing, optimizing, or understanding LimaCharlie Query Language (LCQL) queries for searching telemetry.
ai-partner-chat
基于用户画像和向量化笔记提供个性化对话。当用户需要个性化交流、上下文感知的回应，或希望 AI 记住并引用其之前的想法和笔记时使用。
hierarchical-reasoning
Implements sophisticated multi-level reasoning for complex problems requiring strategic planning, tactical approach design, and operational execution. Use for problems needing deep analysis, multi-step reasoning, systematic decomposition from first principles to implementation, convergence-aware iterative refinement, or uncertainty quantification across abstraction levels.
piper
Convert text to speech using Piper, a fast, local, neural text-to-speech system with natural sounding voices. This skill is triggered when the user says things like "convert text to speech", "text to audio", "read this aloud", "create audio from text", "generate speech from text", "make an audio file from this text", or "use piper TTS".
python-ai-expert
Comprehensive Python AI/ML development expert with 10+ years experience using UV package manager. Covers PyTorch, TensorFlow, scikit-learn, transformers, langchain, pandas, numpy, OpenCV, and all major AI/ML libraries. Automatically audits projects, generates production-ready code with type hints, optimizes performance, sets up RAG pipelines, manages dependencies with UV, and ensures best practices. Use for AI/ML project setup, model training, data processing, LLM applications, computer vision, code generation, dependency management, performance optimization, and comprehensive project auditing. Excludes Tkinter and desktop UI libraries.
notebooklm
Enables Claude to create and interact with NotebookLM for document analysis, audio overviews, and knowledge synthesis
resolving-uncertainty
Turns ambiguity into an Uncertainty Register, prioritizes items, and produces observation tasks/experiments with evidence and decision rules. Use when the user mentions uncertainty/unknowns/assumptions/validation/investigation/hypotheses/risks or cannot decide (不確実性, 曖昧, 未知, 前提, 仮説, 検証, 調査, 優先順位, 観測, 意思決定).
llm-eval-designer
LLM生成システムの検証設計スキル。assay-kitフレームワークを活用し、LLM特有の失敗モード（幻覚、例への過学習、部分的処理）を考慮した総合的なテストケース設計を支援する。 使用タイミング: - LLMベースのワークフロー/エージェントの評価設計時 - ゴールデンデータセット（golden-dataset.yaml）の設計・拡張時 - 既存テストが特定パターンに過学習していないか検証時 - LLM出力の品質スコアラー設計時 - 「なぜこのテストケースが必要か」の根拠を示す時
rules-optimizer
Optimize and create Claude Rules (.claude/rules/*.md) following best practices.
