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 benchmark evaluates Vision-Language Models (VLMs) on detecting potential safety hazards within indoor environments. It uses static camera frames, simulating real-world fixed security camera monitoring. The test covers 40 scenarios across 5 critical categories, including fire/smoke, electrical risks, trips/falls, child safety, and falling objects, providing a rigorous measure of VLM's practical safety inspection capability.

A robust framework for building complex, long-horizon AI agents using LangGraph. It addresses common pitfalls in advanced agent design, such as unbounded state growth (Virtual FS cleanup) and subagent persona leakage (explicit message overriding). Ideal for tasks requiring multi-step planning, task delegation, and self-correction through reflection loops.

Master the complexities of building robust Retrieval Augmented Generation (RAG) pipelines using LangChain. This guide addresses common pitfalls in vector search, such as score semantics differences (L2 vs. Cosine), embedding dimension mismatches, and chunking issues for code/Markdown. Learn best practices for selecting between FAISS, Pinecone, Chroma, and PGVector, normalizing scores, and implementing hybrid keyword+vector search for maximum retrieval quality.

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.

Implement robust human-in-the-loop (HITL) approval gates for sophisticated LangGraph workflows. This skill enables the graph to pause execution before critical, irreversible actions (like sending emails or calling tools). It manages complex state serialization, provides clear resume semantics using `Command(resume=...)`, and integrates with external UIs (e.g., Slack) to await human decision, ensuring safe and controlled workflow execution.

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.

Master advanced patterns for composing LangGraph subgraphs, which is essential for building robust, multi-agent systems like planner-executors or hierarchical teams. This skill addresses critical issues such as shared state contract propagation (P21) and callback scoping (P28). Learn to manage state across independent graph components using sophisticated dispatch patterns: compiled nodes, Send fan-out, and Command bubble-up.

This skill integrates LangChain 1.0 and LangGraph 1.0 into standard OpenTelemetry (OTEL) backends (e.g., Jaeger, Honeycomb, Tempo). It solves critical observability gaps by ensuring correct prompt content capture, propagating spans across complex subgraphs, and providing five LLM-specific Service Level Objectives (SLOs) like latency and cost. Ideal for multi-tenant, compliance-sensitive, or existing OTEL stack environments where LangSmith is unsuitable.

A comprehensive skill for generating high-quality AI images and videos using advanced models like Seedance 2.0, Kling 3.0, and Midjourney. It supports diverse workflows including text-to-image, image-to-video, and video continuation. The system handles asynchronous generation, robust delivery, and multi-channel integration (e.g., Feishu, Slack), making it ideal for creative content production and automated media workflows.

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.