Skill UI

A comprehensive evaluation suite designed to benchmark the performance of Large Language Models (LLM) and Vision-Language Models (VLM) specifically for home security applications. It covers 143 diverse tests across 16 critical domains, including tool use, prompt injection resistance, contextual reasoning, alert triage, and detailed scene analysis, ensuring robustness for smart home AI systems.

This harness provides comprehensive, reproducible evaluation pipelines for complex LLM chains and agents (LangChain/LangGraph 1.0). It integrates golden dataset management, LangSmith evaluation, RAGAS metrics, deepeval LLM-as-judge, and structured agent trajectory analysis. Use it when establishing quality benchmarks for new chains, diagnosing performance regressions after model switches, or implementing CI/CD gates to prevent quality drops.

This guide addresses critical LangGraph streaming pitfalls, including state overdraw, proxy buffering issues (Nginx, Cloud Run), and excessive event noise. It provides a production-grade pattern for implementing robust SSE/WebSocket endpoints using FastAPI. Learn how to correctly select the `stream_mode` (messages, updates, values) and implement server-side event filtering to reliably build live-token chat UIs and perform comprehensive state visualization.

This powerful CLI allows users to interact with the MiniMax AI platform directly from the terminal. It supports generating and managing diverse multi-modal content, including text, images, video, speech, and music, in addition to performing web searches and analyzing images. It is essential for integrating complex AI capabilities into automated workflows, CI/CD pipelines, or agent-based systems.

This comprehensive skill manages the entire computer vision pipeline, from custom dataset annotation to model deployment. Users can fine-tune YOLO models on proprietary COCO-formatted datasets using an agent-driven workflow. Key features include hardware-aware training (supporting CUDA, MPS, ROCm), automatic export to optimized formats (TensorRT, CoreML, OpenVINO, ONNX), and seamless, one-click deployment as an active detection skill.

This is an all-in-one command-line tool for generating and editing images using the full Nano Banana model family. Users can select the optimal model—Gemini 3.1 Flash for balanced performance, Gemini 3 Pro for high-quality output, or Gemini 2.5 Flash for speed—to suit their needs. Key features include text-to-image generation, precise single image editing, and multi-image collage composition, making it versatile for creative and technical workflows.

This skill orchestrates the entire patent drafting lifecycle, guiding users from an initial invention concept to legally formatted filing documents. It manages critical steps such as prior art search, claims design, specification writing, and jurisdiction-specific formatting. Supports Invention and Utility Models across CN, US, and EP jurisdictions.

This skill provides comprehensive analysis and improvement suggestions for plugin prompts, skill definitions, and command instructions. It evaluates content across five dimensions—clarity, safety, effectiveness, completeness, and conciseness—providing a detailed score card and concrete, actionable rewrites to ensure compliance and optimal AI performance.

Qdrant is a high-performance vector database. This client SDK suite provides official, easy-to-use libraries for multiple programming languages (including Python, Java, Go, and TypeScript). It enables developers to seamlessly integrate sophisticated vector search capabilities into their applications, supporting both REST and gRPC APIs. Ideal for building semantic search engines, recommendation systems, and advanced AI applications.

This guide provides detailed strategies for optimizing Qdrant's operational performance. It covers critical areas such as search speed (latency and throughput), efficient vector indexing strategies, and advanced memory management techniques. Use this when you need to improve the speed, scalability, and resource efficiency of your vector database deployment.

A comprehensive guide for diagnosing and resolving slow search performance in Qdrant. This resource covers issues related to low QPS (throughput), high latency for single queries, and slow filtered searches. Solutions include tuning HNSW parameters, managing memory pressure, implementing quantization, optimizing indexing, and configuring cluster settings for maximum efficiency.

This skill integrates with the Runway ML platform, enabling developers to programmatically generate high-quality video and edit images using the official Python SDK. Use this when your workflow requires advanced creative AI features, such as converting text prompts into videos, or managing complex creative tasks within a CI/CD pipeline. The skill provides necessary patterns and error handling for successful API calls.