Skill UI

This skill provides comprehensive commands to manage the entire AI model evaluation lifecycle using Arize. Users can list, retrieve metadata, and export results from experiments. It is ideal for comparing multiple model runs, benchmarking performance against a dataset, and analyzing model accuracy using structured evaluation workflows (CRUD operations).

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.

PyTDC (Therapeutics Data Commons) is an open-science platform providing comprehensive, AI-ready datasets and benchmarks for the entire drug discovery and development pipeline. It supports single-instance prediction (e.g., ADME, toxicity), multi-instance prediction (DTI, DDI, PPI), and novel molecule generation (MolGen). Use this skill for benchmarking machine learning models and predicting complex biomedical properties.

Comprehensive solution for conducting load testing, performance benchmarking, and auto-scaling setup for various Adobe APIs (Firefly, PDF Services, Photoshop). This guide utilizes k6 for stress testing, simulating real-world user traffic, and includes Kubernetes HPA configuration to ensure the application maintains high availability and stability under peak loads.

This guide demonstrates how to deploy GPU workloads on the CoreWeave platform using Kubernetes (kubectl). It covers deploying vLLM inference services for large language model (LLM) testing and running batch CUDA jobs for benchmarking. This is ideal for users running their first GPU workloads, testing inference pipelines, or verifying cluster access on bare-metal GPU nodes (A100, H100).

This guide provides expert strategies for optimizing GPU inference performance on CoreWeave infrastructure. It covers GPU selection based on workload (LLM, image generation, training), advanced techniques like continuous batching (using vLLM), autoscaling setup (HPA), and benchmarking data. Use this to maximize GPU utilization, minimize latency, and optimize large-scale AI model deployment.

This advanced engine creates deeply personalized, relationship-aware daily gifts using a sophisticated five-stage creative pipeline. It first performs an 'Editorial Judgment' to determine if a gift is warranted, and then generates the content concept. Output can be rendered as interactive H5 web pages, AI-generated images, or videos, ensuring that every artifact is highly unique, contextually relevant, and avoids repetition. Ideal for automated journaling, marking milestones, or maintaining thoughtful daily engagement.

A universal infrastructure for automatically evolving and optimizing AI agents across any domain. Using LLM-driven mutation and iterative solve-observe-evolve cycles, it enhances critical components like prompts, skills, and memory by benchmarking performance against measurable datasets, enabling the creation of self-improving AI systems.

Implement robust data governance and security measures within Palantir Foundry. This skill guides users through PII redaction, applying data markings (e.g., CONFIDENTIAL, RESTRICTED), ensuring GDPR/CCPA compliance, and establishing automated data retention policies. It provides methods for column-level security and auditable deletion workflows for sensitive information.

This guide teaches best practices for configuring robust, enterprise-grade Role-Based Access Control (RBAC) within Palantir Foundry. Learn how to define project roles (Viewer, Editor, Owner), manage organization-level groups to enforce least privilege, establish secure service user accounts for external integrations, and utilize marking-based data classification to ensure data governance and security across large datasets.

This tool evaluates how thoroughly a feature implementation meets product requirements (PRDs) or specifications. It scores acceptance criteria case-by-case, independently grading both the code implementation and the associated test coverage (unit and E2E). Ideal for benchmarking multiple development efforts, auditing adherence to specifications, or determining if a feature is 100% complete and compliant with the ground truth PRD.

This skill guides the systematic triage of project issues and external pull requests (PRs). It moves development items through a defined state machine using roles (e.g., bug, enhancement) and states (e.g., ready-for-agent, wontfix). The process involves context gathering, cross-codebase verification, deep questioning (grilling), and generating comprehensive, agent-ready briefs or marking them for human implementation.