What it is: A curated, searchable directory of 157+ community and official skills/plugins for Claude Code, filterable by tag and publisher. Each entry shows a five-point security profile (code execution, data transmission, external API calls, credential access, auto-update behavior) alongside dependency and install info. Contributors range from Anthropic and established vendors to individual developers, so quality varies by entry.
☝️ Opinion: Good listing of skills and shows dependencies and external attributes up front.
What it is: An open-source frontend skill/framework that installs into AI coding agents (Cursor, Claude Code, Codex, Gemini CLI, v0, Lovable) to stop them generating generic, templated UI. It ships design system rules, component guidelines, multiple style variants (minimalist, brutalist, soft), dark mode support, and audit-first pre-flight checks for redesigns.
☝️ Opinion: Content review for editing and voice - needs detailed testing.
What it is: A skill/plugin system for AI coding agents that pushes toward minimal, efficient code generation. It works via a decision ladder the agent checks before writing anything — reuse existing code, use the standard library, use native platform features, use an installed dependency, or write one line — only falling back to custom code when none of those cover it. Works across 20+ AI agents and coding platforms (Claude Code, Cursor, GitHub Copilot CLI, Codex, Devin, and others), and its own benchmarks claim ~54% less generated code on average.
☝️ Opinion: Semi-useful. Has more hooks than I like. I’ve only seen it kick in a few times.