Skill UI

This skill is designed for performing local, on-premise evaluations and benchmarking of models hosted on the Hugging Face Hub. It facilitates running specialized evaluation frameworks like `inspect-ai` and `lighteval` against local hardware. Users can select optimal inference backends, including `vllm` for high throughput, or fall back to Hugging Face Transformers or `accelerate`, ensuring comprehensive smoke testing and task selection for LLMs. It focuses purely on the local execution process, separate from remote jobs or publishing workflows.

This tool measures whether coding agents adhere to predefined skills, rules, or operational definitions. It automatically generates multiple levels of scenario prompts (from supportive to competitive) and analyzes the full tool call timeline. By using LLMs to classify actions, it provides a detailed, self-contained compliance report and score, validating adherence even when the prompt is ambiguous.

Collection of battle-tested PyTorch training recipes covering architecture selection, optimizers, LR scheduling, precision tips, and debugging to train or fine-tune LLMs, vision, diffusion, and biomedical models with high GPU throughput.

Covers every key Golang documentation surface—godoc comments, README, CONTRIBUTING, CHANGELOG, Example tests, Go Playground demos, API docs, llms.txt, and doc-site guidance—so you can write, review, and maintain discoverable docs for both libraries and CLIs.

Behavioral guidelines for writing, reviewing, and refactoring code with LLMs, emphasizing minimalism, clear assumptions, surgical edits, and verifiable success criteria to avoid common machine-assisted coding mistakes.

This skill provides comprehensive management capabilities for AI Integrations within the Arize platform. Users can create, list, retrieve, and update credentials for various Large Language Model (LLM) providers, including OpenAI, Anthropic, Azure OpenAI, AWS Bedrock, and Vertex AI. These integrations are crucial for enabling Arize's advanced monitoring and evaluation features that require invoking external LLMs.

Describes how to analyze a repository, prioritize key docs, and author an llms.txt entry so LLMs immediately understand the project scope, documentation clusters, and configuration artifacts they should consult.

This skill guides the end-to-end process of building automated evaluation pipelines for Python applications powered by Large Language Models (LLMs). It teaches how to define evaluation criteria, instrument the application, build golden datasets, and run real-world evaluations using tools like pixie, ensuring the app's logic (not just the LLM) is rigorously tested. Ideal for QA, benchmarking, and improving LLM-based services.

A comprehensive guide to optimizing website content for modern AI search systems, including ChatGPT, Perplexity, and Google AI Overviews. It covers structural improvements, maximizing citability scores, incorporating multi-modal elements, building brand authority, and implementing technical standards like llms.txt and proper server-side rendering (SSR). Essential for maximizing organic reach in the LLM era.

Maintains the repository's llms.txt guidance file by auditing current documentation, specs, and structure to ensure the navigation summary matches reality and stays compliant with the official specification.

This guide covers the entire South Asian film ecosystem, including Bollywood (Hindi), Tollywood (Telugu), Kollywood (Tamil), Malayalam, Bengali, and Pakistani cinema. It teaches users how to identify regional clues and conduct comprehensive searches, ensuring results are not limited only to Hindi films.

This technique is designed to identify South Asian (Indian/Bollywood) films when the question provides insufficient information to name the movie directly. Instead, it requires chaining together the biographical details of the actor or director—such as birth year, debut era, family relations, or awards—to first uniquely identify the person, and then use that identity to locate the target film and verify all constraints.