设计 AI 架构，编写 Prompt，构建 RAG 系统和 LangChain 应用

ai-llm-engineering

Operational patterns for LLM inference: latency budgeting, tail-latency control, caching, batching/scheduling, quantization/compression, parallelism, and reliable serving at scale. Emphasizes production-grade performance, cost control, and observability.

Guide for AI Agents and LLM development skills including RAG, multi-agent systems, prompt engineering, memory systems, and context engineering.

Production LLM engineering skill. Covers strategy selection (prompting vs RAG vs fine-tuning), dataset design, PEFT/LoRA, evaluation workflows, deployment handoff to inference serving, and lifecycle operations with cost/safety controls.

Manage AI agents through the AI Maestro CLI. This skill provides commands for creating, updating, deleting, hibernating, and waking agents. It also handles plugin management and agent import/export.

AI-powered marketing engineering skill based on Alon Huri's framework. Transforms marketing from copywriting to engineering discipline through 10 agentic mechanisms: infinite creative generation, adaptive budget management, LTV signal hunting, contextual data layers, AEO optimization, dynamic quizzes, behavior-driven activation, personalized video at scale, competitor weakness targeting, and active churn prevention. Use when building marketing automation systems, designing growth engineering workflows, creating AI-powered marketing agents, optimizing ad creatives at scale, implementing AEO (Answer Engine Optimization), or architecting data-driven marketing infrastructure.

End-to-end data science and ML engineering workflows: problem framing, data/EDA, feature engineering (feature stores), modelling, evaluation/reporting, plus SQL transformations with SQLMesh. Use for dataset exploration, feature design, model selection, metrics and slice analysis, model cards/eval reports, experiment reproducibility, and production handoff (monitoring and retraining).

AI and machine learning development with PyTorch, TensorFlow, and LLM integration. Use when building ML models, training pipelines, fine-tuning LLMs, or implementing AI features.

ai-ml-engineer

AI and ML expert including PyTorch, LangChain, LLM integration, and scientific computing

AI/ML APIs, LLM integration, and intelligent application patterns

Operational patterns, templates, and decision rules for time series forecasting (modern best practices): tree-based methods (LightGBM), deep learning (Transformers, RNNs), future-guided learning, temporal validation, feature engineering, generative TS (Chronos), and production deployment. Emphasizes explainability, long-term dependency handling, and adaptive forecasting.

Production MLOps and ML/LLM/agent security skill for deploying and operating ML systems in production (registry + CI/CD, serving, monitoring/drift, evaluation loops, incident response/runbooks, and governance), including GenAI security (prompt injection, jailbreaks, RAG security, privacy, and supply chain).

A production-ready pattern for integrating AI models (specifically Google Gemini) with automatic fallback, retry logic, structured output via Zod schemas, and comprehensive error handling. Use when integrating AI/LLM APIs, need automatic fallback when models are overloaded, want type-safe structured responses, or building features requiring reliable AI generation.

Complete guide for calling AI models with CloudBase - covers JS/Node SDK and WeChat Mini Program. Text generation, streaming, and image generation.

AI 모델 API 호출명 및 가격 참조 가이드. API 키로 AI 모델을 호출할 때 정확한 모델명(model string)과 최신 가격 정보를 제공합니다. 사용 시점: (1) OpenAI, Anthropic, Google, DeepSeek 등의 API 호출 시 모델명이 필요할 때, (2) 토큰 비용/가격 비교가 필요할 때, (3) 최신 추론 모델/FAST 모델/가성비 모델 선택이 필요할 때, (4) 프롬프트 캐싱/배치 처리 비용 최적화가 필요할 때

Gère les modèles IA de Motivia. Utilise ce skill quand l'utilisateur demande d'ajouter un nouveau modèle IA, modifier un provider, ou configurer les options de génération. Supporte OpenAI, Anthropic, Google, Mistral et xAI.

Process and generate multimedia content using Google Gemini API for better vision capabilities. Capabilities include analyze audio files (transcription with timestamps, summarization, speech understanding, music/sound analysis up to 9.5 hours), understand images (better image analysis than Claude models, captioning, reasoning, object detection, design extraction, OCR, visual Q&A, segmentation, handle multiple images), process videos (scene detection, Q&A, temporal analysis, YouTube URLs, up to 6 hours), extract from documents (PDF tables, forms, charts, diagrams, multi-page), generate images (text-to-image with Imagen 4, editing, composition, refinement), generate videos (text-to-video with Veo 3, 8-second clips with native audio). Use when working with audio/video files, analyzing images or screenshots (instead of default vision capabilities of Claude, only fallback to Claude's vision capabilities if needed), processing PDF documents, extracting structured data from media, creating images/videos from text prompts, or implementing multimodal AI features. Supports Gemini 3/2.5, Imagen 4, and Veo 3 models with context windows up to 2M tokens.

Multimodal AI processing via Google Gemini API (2M tokens context). Capabilities: audio (transcription, 9.5hr max, summarization, music analysis), images (captioning, OCR, object detection, segmentation, visual Q&A), video (scene detection, 6hr max, YouTube URLs, temporal analysis), documents (PDF extraction, tables, forms, charts), image generation (text-to-image, editing). Actions: transcribe, analyze, extract, caption, detect, segment, generate from media. Keywords: Gemini API, audio transcription, image captioning, OCR, object detection, video analysis, PDF extraction, text-to-image, multimodal, speech recognition, visual Q&A, scene detection, YouTube transcription, table extraction, form processing, image generation, Imagen. Use when: transcribing audio/video, analyzing images/screenshots, extracting data from PDFs, processing YouTube videos, generating images from text, implementing multimodal AI features.

A complete end-to-end framework for non-technical product managers to build and ship software using AI coding agents. Use this when starting a side project, building a prototype, or automating internal tools without an engineering team.

Proactively manage development tasks in TASKS.md. Automatically tracks progress, updates status, prioritizes backlog, and estimates effort. Runs in background during development - no explicit invocation needed.

Build AI-first applications with RAG pipelines, embeddings, vector databases, agentic workflows, and LLM integration. Master prompt engineering, function calling, streaming responses, and cost optimization for 2025+ AI development.

AI-native product building shifts the Product Manager’s role from writing requirements for others to directly directing an AI agent to build the software. This approach reduces development time from m

In the AI era, product market fit must be constantly "refounded." This framework moves teams away from "blunt instruments" (long roadmaps, rigid PRDs) toward a high-velocity, hands-on approach where t

Crawl the latest AI-related news from multiple websites, merge and deduplicate them, select 10 in descending order of time, unify summaries into Chinese, and write to a JSON file.

>

Manage AI notes in ~/Compost/AI-Notes - read existing notes, list recent documents, and write new implementation plans (仕様書, 計画書, 設計書). Auto-triggers when reading from or writing to notes, specifications, or plans.

Design AI-powered immersive NPC systems for escape room games using proven actor techniques from Korean immersive escape rooms (Danpyeonsun, Ledasquare). Implements adaptive dialogue, emotional simulation, player profiling, and trust dynamics using Gemini/GPT-4. Creates character profiles with lying probabilities, improvisational responses, and cost-optimized streaming. Use for murder mystery NPCs, suspect interrogation, or dynamic character interactions.

Implement or tune AI opponent behavior

Multi-AI engineering loop orchestrating Claude, Codex, and Gemini for comprehensive validation. USE WHEN (1) mission-critical features requiring multi-perspective validation, (2) complex architectural decisions needing diverse AI viewpoints, (3) security-sensitive code requiring deep analysis, (4) user explicitly requests multi-AI review or triple-AI loop. DO NOT USE for simple features or single-file changes. MODES - Triple-AI (full coverage), Dual-AI Codex-Claude (security/logic), Dual-AI Gemini-Claude (UX/creativity).

Multi-model AI collaboration via orchestrator MCP. Use when seeking second opinions, debugging complex issues, building consensus on architectural decisions, conducting code reviews, or needing external validation on analysis.

Build a unified, ADHD-friendly web UI that consolidates 70+ CLI tools into a beautiful liquid glass interface for the AI File Organizer. Use when creating the complete frontend application that replaces all terminal interactions with a macOS-inspired dashboard for file organization, search, VEO prompts, and system management.

ai-output-validation

name: ai-pair-programming

Detects AI-generated writing patterns and suggests authentic alternatives. Auto-applies when reviewing content, editing documents, generating text, or when user mentions writing quality, AI detection, authenticity, or natural voice.

Use when analyzing stakeholder psychology for negotiations, proposals, or persuasion. Creates research-backed personas revealing hidden motivations.

Strategic advisor for founders. Facilitates deep thinking through Socratic dialogue, identifies blind spots, assesses risks and opportunities, and provides guidance on strategic decisions using Extended Thinking for complex analysis.

Protocolo de autodiagnostico contra os 5 problemas mais comuns da IA ao programar. Detecta overengineering, codigo duplicado, reinvencao da roda, falta de documentacao e arquivos monoliticos. Use SEMPRE antes de implementar, ao planejar mudancas, quando criar funcoes novas, ao escrever codigo, para revisar implementacoes. Palavras-chave - simples, duplicado, repetido, existe, separar, modular, documentacao, complexo, refatorar, engenharia demais, roda, reutilizar.

description: Maintain and update AI coding procedures when new elements are added

In the era of LLMs, product development moves from writing static specifications to defining "correctness" through Evals. Since models are stochastic, you cannot "fix a bug" with a single line of code

Every product will be AI-powered. The question is whether you'll build it right or ship a demo that falls apart in production. This skill covers LLM integration patterns, RAG architecture, prompt engineering that scales, AI UX that users trust, and cost optimization that doesn't bankrupt you. Use when: keywords, file_patterns, code_patterns.

A framework to assess and integrate AI into your product strategy by mapping core customer problems to AI capabilities. Use this when your industry is facing a major technology shift, when prioritizing an AI roadmap, or when deciding between augmenting existing features vs. building new AI-first solutions.

Help users define AI product strategy. Use when someone is building an AI product, deciding where to apply AI in their product, planning an AI roadmap, evaluating build vs buy for AI capabilities, or figuring out how to integrate AI into existing products.

Create an AI Product Strategy Pack (thesis, prioritized use cases, system plan, eval + learning plan, agentic safety plan, roadmap). Use for AI product strategy, LLM/agent strategy, AI roadmap, AI-first product direction.

AI engineering skill for prompt optimization, context inference, and intelligent command routing across different models and use cases

Operational prompt engineering for production LLM apps: structured outputs (JSON/schema), deterministic extractors, RAG grounding/citations, tool/agent workflows, prompt safety (injection/exfiltration), and prompt evaluation/regression testing. Use when designing, debugging, or standardizing prompts for Codex CLI, Claude Code, and OpenAI/Anthropic/Gemini APIs.

Expert assistant for managing AI prompts, features, and configuration in the KR92 Bible Voice AI system. Use when creating AI prompts, configuring AI features, managing prompt versions, setting up AI bindings, or working with AI pricing and models. Supports multiple vendors and models for feature flexibility.

Expert assistant for managing AI prompts, features, and configuration in the KR92 Bible Voice AI system. Use when creating AI prompts, configuring AI features, managing prompt versions, setting up AI bindings, or working with AI pricing and models.