Skill UI

This tool extracts and formalizes the core business terminology (Ubiquitous Language) from a conversational context. It generates a structured glossary, identifies ambiguities, resolves conflicting terms, and proposes canonical names. It is essential for building consistent and unambiguous domain models, particularly when following Domain-Driven Design (DDD).

unslop is a powerful command-line interface (CLI) tool designed for post-processing text. It programmatically strips out common and identifiable patterns associated with AI-generated writing. Use it as a final, deterministic pass before publishing documentation, blog posts, or any content derived from large language models (LLMs), ensuring a more natural and human-like tone.

This tool searches a comprehensive database of over 2500 curated open-source repositories for tools related to Large Language Models (LLMs), ChatGPT, RAG, agents, and NLP. It interprets natural language queries, converts non-English inputs to core English keywords, and systematically filters results across multiple specialized categories (e.g., LangChain, Chatbots, CLIs). Use it whenever the user needs to find specific AI development libraries or open-source AI tooling.

This skill helps discover new Japanese Natural Language Processing (NLP) resources, including GitHub repositories and Hugging Face models/datasets, that are not yet documented in the awesome-japanese-nlp-resources list. It analyzes a given topic, suggests candidates using web search, and outputs contribution-ready markdown for easy integration.

A specialized skill for optimizing content to be cited as authoritative sources by major AI language models (ChatGPT, Perplexity, Claude, Gemini). Unlike traditional SEO, AEO focuses on E-E-A-T signals, factual density, and structured data to ensure LLM integration. Features include content auditing, E-E-A-T scoring, structural rewriting, and citation tracking.

A comprehensive tool for generating high-quality audio content, including instrumental background music (BGM), jingles, ambient beds, and full songs with sung vocals. It supports complex creative parameters such as mood, tempo, vocal style, and target language, ensuring original and copyright-safe output through specialized generation models.

Provides expert guidance for authoring and editing Tabular Model Definition Language (TMDL) files directly within PBIP projects. This skill is designed for advanced users needing to perform quick, text-level modifications—such as updating descriptions, fixing format strings, or defining measures—when full, dedicated tools (like Tabular Editor CLI) are unavailable. It details critical syntax rules, object nesting, and file structures for managing complex semantic models.

This skill implements the LLM Council concept, enabling multiple open-weight language models to collaborate on complex queries. The process follows a three-stage deliberation: 1) Parallel independent responses from selected models; 2) Peer-to-peer ranking of the responses; and 3) Final synthesis by a designated Chairman LLM. Powered by Fireworks AI, it ensures deep, multi-faceted analysis and consensus building for high-stakes decision support.

NVIDIA garak is an open-source LLM vulnerability scanner designed for comprehensive red-teaming and security assessment. It systematically probes large language models (via OpenAI-compatible APIs, Hugging Face, etc.) using thousands of adversarial prompts. It covers critical vulnerabilities such as prompt injection, jailbreaking, data leakage, and toxicity. Use this tool to establish a repeatable security baseline before or after deployment, validate the effectiveness of safety guardrails, or generate quantitative evidence for AI risk assessments.

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.

A comprehensive guide for training or fine-tuning advanced sentence-transformer models. This skill assists in selecting the correct architecture—including Bi-encoders (SentenceTransformer), CrossEncoders (for reranking), and SparseEncoders—for various Natural Language Processing (NLP) tasks. Use cases include semantic retrieval, similarity matching, document clustering, and pair-wise classification.

This skill provides expert capabilities for training and fine-tuning transformer language models using the TRL (Transformers Reinforcement Learning) library. It supports state-of-the-art post-training methods, including Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), Group Relative Policy Optimization (GRPO), and Reward Model training. Use this skill to align and customize foundation models for advanced tasks via CLI commands.