Skill UI

This skill guides the design and implementation of a complete, production-grade Machine Learning pipeline. It employs a multi-agent orchestration workflow, systematically covering all stages of the ML lifecycle: data ingestion and quality assurance, feature engineering, model training, optimization, and robust deployment (MLOps). It ensures built-in reproducibility, monitoring, and scalability.

This skill provides comprehensive guidance for building production-grade MLOps pipelines. It covers the entire machine learning lifecycle, from initial data ingestion and rigorous feature engineering to model training, advanced validation (A/B testing, performance drift), and robust deployment strategies (Canary, Blue-green). It enables the design and orchestration of scalable, reproducible, and observable ML systems using tools like Airflow and Kubeflow.

A comprehensive framework designed to structure and automate PyTorch deep learning workflows. It simplifies boilerplate code by organizing models into LightningModules, automating training loops via the Trainer, and providing robust support for multi-GPU/TPU scaling (DDP, FSDP, DeepSpeed), data handling (DataModules), and best-in-class experiment logging (W&B, MLflow). Ideal for professional, scalable deep learning project development.

This reference architecture provides a comprehensive blueprint for deploying Machine Learning models on CoreWeave's GPU cloud infrastructure. It details the structure for multi-model serving (e.g., vLLM, TGI), scalable Kubernetes deployments, shared persistent storage (PVC), and autoscaling mechanisms using KServe/Knative. Use this when designing robust MLOps pipelines, planning high-performance multi-model inference services, or establishing standardized best practices for GPU cloud deployment.

A robust framework for building complex, long-horizon AI agents using LangGraph. It addresses common pitfalls in advanced agent design, such as unbounded state growth (Virtual FS cleanup) and subagent persona leakage (explicit message overriding). Ideal for tasks requiring multi-step planning, task delegation, and self-correction through reflection loops.

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.

The /goal command converts standard agent prompts into persistent, self-correcting loops (plan → act → test → review → iterate). It is designed for long-running, complex engineering tasks like migrations, coverage lifts, or refactoring, ensuring progress is tracked against a verifiable stop condition and mandated documentation.

A powerful tool designed to scaffold sophisticated, custom AI agent harnesses using the metaharness CLI. It supports various templates (e.g., coding, devops, marketing) and AI hosts (e.g., Claude, Codex). Crucially, it operates with maximum safety, defaulting to a dry-run mode and refusing to write to the calling repository root, ensuring controlled and predictable development workflows.