Skill UI

Comprehensive guide to LLM Operations (LLMOps). Covers advanced topics like Retrieval-Augmented Generation (RAG), vector database management (Pinecone, Chroma), embeddings, fine-tuning, prompt engineering, and cost optimization. Essential for bridging the gap between AI prototypes and reliable, scalable, production-grade AI systems.

This skill provides comprehensive techniques for optimizing prompts used with any Large Language Model (LLM). It teaches proven methodologies—including Role definition (R), Context setting (S), Constraints (C), Instructions (I), and Templates (T)—to transform vague inputs into precision-engineered instructions. Users can reliably achieve structured JSON output, implement Chain-of-Thought reasoning, reduce hallucinations, and minimize token usage, ensuring consistent and high-quality results across different AI platforms.

Conduct authorized security assessments for LLM applications and AI agents. Covers prompt injection, tool abuse, RAG exposure, memory poisoning, and system prompt extraction following OWASP LLM/ASI Top 10 guidelines. Includes red-teaming, vulnerability mapping, and compliance engineering for AI systems to ensure safety and robustness.

This skill provides methods for reliably extracting typed, validated, and structured data from Large Language Models (LLMs), moving beyond simple text parsing. It covers best practices across major APIs (OpenAI's response_format, Anthropic's tool_use, Google's responseSchema) and local models. Key workflows include defining schemas (Pydantic/Zod), implementing constrained decoding, and building robust retry loops for production-grade data pipelines.

Design and operate production LLMOps platforms with CI/CD, model promotion workflows, evaluation gates, rollback, and governance for cloud and self-hosted inference.

An expert guide for deploying and optimizing Large Language Models (LLMs) entirely on local hardware. Mastering the entire local AI ecosystem, including inference engines like Ollama, vLLM, and llama.cpp. Provides deep knowledge on quantization formats (GGUF, EXL2, AWQ), VRAM calculation, and ensuring privacy-first, offline deployment for state-of-the-art models.

Configure a Mac mini as a reliable local LLM server with remote access, observability, and power-safe operation. Use when building an always-on private AI inference server on Apple Silicon.

Build Model Context Protocol servers to enable LLMs to interact with external APIs and services. Supports Python (FastMCP) and Node/TypeScript (MCP SDK) for creating tools that help AI agents accomplish real-world tasks. Learn to design well-structured tools, manage context, and implement transport mechanisms like stdio or Streamable HTTP.

MockHunter is an advanced skill designed to audit live web pages, providing a comprehensive reality check on visible data. It analyzes every element to determine its true source: whether it is REAL (from the database), MOCK (simulated data), LLM-generated, HARDCODED, BROKEN (broken endpoints), or UNKNOWN. Ideal for auditing AI-generated UIs or reviewing proof-of-concept MVPs before customer sign-off.

Best practices for building AI chat interfaces with Makepad using MolyKit, covering cross-platform async, OpenAI/LLM integration, SSE streaming, chat widgets, and bot clients.

Design secure multi-tenant LLM hosting platforms on shared GPU inference infrastructure, covering tenant isolation, per-tenant quotas and rate limits, billing attribution, noisy-neighbor protection, and per-tenant policy controls for vLLM on Kubernetes.

Unified LLM inference layer built into Neon branches: one credential and one OpenAI-compatible API for frontier and open-source models from seven providers. Switch models without rewiring, wire it into OpenAI/Anthropic SDKs, and keep AI requests branch-scoped.