Skill UI

Provides calibration-free weight quantization for LLMs down to 1-bit precision with multiple optimized backends, ideal for fast deployments with vLLM/HuggingFace and LoRA/PEFT fine-tuning without collecting calibration data.

Outlines guarantees valid JSON/XML/code outputs by applying CFG/FSM-constrained token filtering, Pydantic schemas, and grammar-level sampling while supporting local Transformers, llama.cpp, and vLLM backends for fast inference.

Guides training large language models with Volcano Engine's verl library, covering PPO/GRPO workflows, distributed backends, multi-turn rollouts, reward configs, critic-enabled PPO, and production deployments using Ray workers.

This skill provides comprehensive tools for documenting and managing model performance on Hugging Face. It supports multiple methods for adding structured evaluation data, including extracting tables from README content, importing benchmark scores via APIs (like Artificial Analysis), and executing high-performance custom evaluations using advanced backends such as vLLM or lighteval. It ensures that all metadata is structured and compatible with model leaderboards and industry standards.

This skill is designed for running comprehensive evaluations of Hugging Face Hub models on local hardware. It supports multiple evaluation frameworks, including `inspect-ai` and `lighteval`, and provides flexibility in backend selection (e.g., `vllm`, Transformers, or `accelerate`) to optimize for performance and compatibility. Use this for local GPU benchmarking, smoke testing, and backend selection, but note that it does not handle remote job orchestration or result publishing.

Transformers.js allows developers to run state-of-the-art machine learning models directly within JavaScript/TypeScript environments (browser, Node.js). It supports a comprehensive range of tasks including NLP (text generation, summarization, translation), Computer Vision (image classification, object detection), and Audio processing. This capability enables building fully client-side, multimodal AI applications without requiring a dedicated Python backend or server infrastructure.

This skill guides the generation of a complete, production-ready Model Context Protocol (MCP) server using C# and .NET. It covers advanced architectural patterns, including Host builder setup, proper dependency injection, structured logging, and implementing robust functional tools. Ideal for developers building sophisticated AI agents or backend services that require standardized communication and tool invocation capabilities.

This comprehensive guide details how to construct a complete, robust Model Context Protocol (MCP) server using TypeScript and the specialized v2 SDK. It covers all aspects of Node.js project setup, selecting appropriate transport layers (HTTP or stdio), implementing complex, schema-validated tools, and integrating resources, resulting in a highly type-safe AI backend service.

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.

Generates high-quality, short video clips (3-15s) using the Seedance 2.0 model. It accepts structured English prompts, an optional starting frame image, and style/identity reference images to ensure character and visual continuity. Supports multiple cloud backends (OpenRouter, Volcengine, BytePlus), making it ideal for storyboarding, cinematic concept visualization, or short film pre-production.

SkillOpt-Sleep provides a mechanism for autonomous agent self-improvement. By simulating a 'sleep' or 'dream' cycle, it harvests past interaction sessions, mines recurring tasks, and replays them through the designated backend. It consolidates learned knowledge into persistent memory and skills, ensuring that only strictly validated improvements are adopted via a held-out gate, simulating long-term competence gain from short-term experience. Ideal for scheduled, background optimization.

A comprehensive guide and skill for deploying large language models (LLMs) locally using the llama.cpp framework and GGUF quantization format. It covers searching the Hugging Face Hub, selecting optimal quantizations (e.g., Q4_K_M), running local servers, and achieving OpenAI-compatible API endpoints. Supports various hardware backends including CPU, Mac Metal, CUDA, and ROCm. Ideal for developers needing offline, private LLM inference.