advanced-evaluation

Filesystem-based context engineering skill for offloading agent context to files, enabling dynamic context discovery, just-in-time context loading, and effectively unlimited memory through file-based storage and retrieval patterns

Designs distinctive, production-grade frontend interfaces with creative aesthetics, generating polished web components and UI that avoid generic AI looks.

Generates comprehensive UI/UX design recommendations with 50+ styles, 21 color palettes, font pairings, and best practices across multiple tech stacks.

Educational skill covering context window fundamentals, attention mechanics, progressive disclosure, and context optimization strategies for AI agent architecture.

Generates unique design philosophies and transforms them into visually stunning artworks in .png and .pdf formats, emphasizing original artistic expression.

Generates structured markdown files for complex task planning, tracking progress, and automatically managing multi-step workflows with session recovery.

A Claude Code skill for LLM project development methodology, covering task-model fit evaluation, batch pipeline design, agent architecture, and cost estimation.

A Claude Code skill for leveraging superpowers capabilities, part of an antigravity skills collection focused on LLM project development and agent architectures.

A Claude Code skill for systematic debugging approaches, part of an antigravity skills collection covering LLM project development and agent-assisted workflows.

Designs and coordinates multi-agent systems by distributing complex tasks across specialized agents with isolated contexts for enhanced problem-solving capabilities.