A 6-skill engineering toolkit that runs natively across Claude Code, Gemini CLI, and OpenAI Codex. Design principle: telegraph style, opinionated, zero filler. Includes debug-hypothesis (scientific debugging), spec-driven-dev (6-gate development), wiki-aggregate (multi-source research with citation tracking), and nano-banana (Gemini image generation). MIT licensed.
A Claude Code Skill with 13 editorial-quality diagram types, each offering minimal-light, minimal-dark, and full-editorial visual variants. Outputs self-contained HTML+SVG with zero dependencies and zero build steps. Automatically extracts brand color systems from a website URL, with WCAG AA contrast verification. 1,127 stars, MIT licensed.
A Claude Code Skill that turns Claude into an AI designer with knowledge of 20+ design philosophies and 68+ design systems. Generates high-fidelity HTML prototypes with interactive components and MP4 video output. Built-in brand style cloning and a five-dimension design critique framework. 544 stars, MIT licensed.
A lightweight React component that renders an animated light-beam border effect around any child element. Implemented with CSS @property for GPU-accelerated animation with no JavaScript animation loop. Auto-detects dark/light theme, 20+ customizable props. 566 stars in 9 days, MIT licensed.
Google's open-source AI file type detection — the system behind Gmail, Drive, Safe Browsing, and VirusTotal. Uses deep learning to read full file content patterns instead of magic bytes, catching disguised malicious files. ~99% accuracy across 200+ formats, ~5ms per file on CPU, multi-language SDK (Python, Rust, JS, Go). Main limitation: encrypted/compressed files get low confidence, and it's ~100-1000x slower than `file`.
A Claude Code skill that turns natural language descriptions into publication-ready SVG and 1920px PNG diagrams. Offers 7 visual styles — including Dark Terminal, Blueprint, and Glassmorphism — and 14 diagram types with full UML coverage. Worth it for engineers who need architecture diagrams that don't look like they were made in MS Paint. Main catch: requires rsvg-convert for PNG export, and auto-layout breaks down past 30 nodes.