Skill UI

Conducts comprehensive AI prompt safety reviews that analyze bias, misinformation, injection, and privacy risks while delivering effectiveness evaluations, security guidance, and best-practice recommendations for prompt engineers.

This skill provides comprehensive commands for creating, managing, versioning, and querying structured datasets within the Arize platform. It covers dataset CRUD operations, appending examples, and exporting data in various formats (JSON, CSV, Parquet) using the `ax` CLI. Ideal for preparing test data, generating evaluation examples, or retrieving golden datasets for LLM development and model testing.

This skill guides users on setting up and running LLM-as-judge evaluations using Arize. It covers defining evaluators (including prompts, classification choices, and models) and executing tasks against real data spans or experiments. It supports key metrics like hallucination, faithfulness, and correctness, and facilitates continuous monitoring for comprehensive LLM performance analysis.

This skill provides a comprehensive framework for optimizing, improving, and debugging Large Language Model (LLM) prompts. It leverages production trace data, system evaluations, and human annotations to extract prompts from spans. Users can run a data-driven optimization loop, ensuring the AI responses meet desired quality standards. Ideal for prompt tuning, system prompt refinement, and improving overall AI output quality.

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.

This system treats ideas like living species on an 'evolution island.' Instead of simple storage, it runs structured, multi-round iterations simulating biological processes: competition, crossbreeding, and mutation. Users input initial concepts, and the tool systematically evaluates them using a 6-dimensional scoring framework (Novelty, Feasibility, Value, etc.). It helps users deepen viability, discover unexpected connections between disparate ideas, and prioritize the most robust concepts for development and investment.

The Phoenix CLI is a powerful tool for debugging complex LLM and agent workflows. It allows users to fetch detailed traces, spans, and sessions, analyze errors, and structure findings using open and axial coding techniques. It is essential for investigating complex LLM failures, building robust failure taxonomies, and guiding the next steps in application evaluation after initial instrumentation.

Phoenix Evals enables developers to build and run evaluators for AI and LLM applications. It supports both code-based and LLM-based evaluation methods, including RAG systems. Users can create datasets, run experiments, validate evaluator accuracy, and implement production guardrails. Available in Python and TypeScript, it helps ensure model performance and reliability through systematic testing and human validation.

A comprehensive expert prompt for evaluating Power BI data models. It guides the assessment of model architecture (Star Schema, relationships), storage optimization (Import, DirectQuery, Dual), data quality, performance scalability, and governance best practices. Use this when reviewing an existing BI model to ensure it is robust, maintainable, and highly performant.

This skill provides a comprehensive audit of user interfaces, evaluating screens against Nielsen's 10 Usability Heuristics and modern mobile UX best practices. It systematically checks for critical usability flaws, design inconsistencies, accessibility gaps, and potential dark patterns, ensuring the resulting product delivers a superior and intuitive user experience across all platforms.

This skill helps users maintain a structured and comprehensive record of their daily activities, progress, key decisions, and learning outcomes. It creates a dedicated file for each day using the ISO date format (yyyy-mm-dd), ensuring that all entries are organized chronologically. By adhering to concise, third-person reporting, it helps users build a valuable, reviewable knowledge base of their journey and accomplishments.

This technique allows users to locate game developers or projects indirectly by searching for highly distinctive biographical details. Instead of relying on game databases, it leverages unique phrases, historical events, or personal lifestyle details found in interviews, blogs, or conference talks. It is invaluable when the developer's identity is obscured, requiring deep research using exact-match queries on obscure information.