Skill UI

Production-ready reinforcement learning library with PPO, SAC, DQN, TD3, DDPG, A2C algorithms. Use for training RL agents, creating custom environments, implementing callbacks, and optimizing training workflows. Best for single-agent RL with Gymnasium environments. Requires Python 3.10+ and PyTorch >= 2.3.

PyTorch Geometric is a standard library for Graph Neural Networks built on PyTorch. It supports node, link, and graph classification, message passing algorithms like GCN and GAT, heterogeneous graphs, and neighbor sampling. Ideal for building scalable GNN models with custom datasets and transforms.

TorchDrug is a modular PyTorch library for graph learning tasks including molecular property prediction, molecule generation, and knowledge graph reasoning. It provides datasets, models, and training engines for building workflows in computational chemistry and biology. Ensure compatibility with PyTorch 1.8-2.0 and Python 3.7-3.10 before implementation.

This pattern library provides production-ready, structured patterns for integrating with the Abridge clinical AI API. It establishes type-safe client wrappers, implements mandatory HIPAA-compliant audit logging (ensuring no Protected Health Information, or PHI, is logged), and incorporates robust error handling with exponential backoff retries. Use this when building secure, scalable, and standards-compliant healthcare AI applications.

This SDK provides production-ready Python patterns for managing GPU workloads on CoreWeave. It includes helpers for defining specific GPU affinity, setting resource limits (A100, H100), and generating complete Kubernetes Deployment YAML. Use this library for building robust inference clients, managing GPU deployments programmatically, or automating complex AI workflows.

The Ejentum Reasoning Harness is a comprehensive library of cognitive operations designed to guide complex AI agents through critical thinking tasks. It exposes four specialized harnesses (reasoning, code, anti-deception, memory) that mitigate common failure modes like hallucination and reasoning decay. Instead of a fixed answer, the agent ingests a detailed, executable scaffold—including failure patterns, procedures, and falsification tests—to ensure robust, self-verified output.

This skill assists users in defining, adapting, and designing bounded, structured feedback loops for AI agents. Whether you need to find a suitable existing workflow, fine-tune a published loop, or design a new process from scratch, this tool guides you through a safe, iterative methodology. It emphasizes setting clear stopping rules, guardrails, and verifiable, measurable steps to ensure the automation remains controllable and focused.

This skill guides the entire lifecycle of repeatable AI agent loops. Use it to discover recurring work patterns within codebases, find established solutions in a library, or interview users to craft bounded, effective loops. It supports auditing, repairing, running, and preparing AI workflows for publication, treating the loop as a robust feedback system.

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.

Scikit-learn is the industry-standard Python library for classical machine learning. It provides comprehensive tools for supervised and unsupervised learning tasks, including classification, regression, clustering, and dimensionality reduction. Use it to build robust, production-ready ML pipelines, perform rigorous model evaluation, and preprocess complex datasets for various data science applications.