Skill UI

This skill covers managing block storage volumes including LVM, cloud-based EBS, filesystem creation, snapshots, and RAID configurations. It guides users through provisioning raw disks, partitioning, formatting, and extending volumes in production environments. Prerequisites include root access and relevant packages like lvm2 and mdadm.

Context optimization extends effective capacity of limited context windows through strategic compression, masking, caching, and partitioning. The goal is to make better use of available capacity rather than increasing window size. Techniques include compaction, observation masking, KV-cache optimization, and context partitioning to reduce costs and latency in long-running agent systems.

Learn to operate GPU-backed Kubernetes clusters for AI inference and training workloads. Covers NVIDIA GPU Operator installation, device plugin configuration, MIG partitioning for resource sharing, and time-slicing for non-MIG GPUs. Includes GPU-aware autoscaling, health monitoring with DCGM and Prometheus, and troubleshooting scheduling or driver issues. Essential for production AI infrastructure management.

Provision, configure, and monitor NVIDIA GPU servers for AI inference and training workloads, including driver/CUDA installation, Docker GPU support, MIG partitioning, and health monitoring.

Design efficient PostgreSQL schemas using best practices for data types, indexing, and constraints. Covers normalization, partitioning, and performance optimization for scalable database systems.

Optimize Apache Spark jobs with partitioning, caching, shuffle optimization, and memory tuning. Covers execution model, performance factors, and practical patterns for improving job speed and scalability.