Building to learn and usually abandon
- inference-lab - Run a bunch of MLX Inference Engines
- guide-mark-ii - Benchmark LLMs on Apple Silicon
- zugar - DSPy-inspired Zod Schema -> prompt + structured output
- garkbit - Automated image generation pipeline powered by local llms
- zod-inquirer - Generate validated inquirer.js prompts from Zod schemas
- fenchurch - Initialize projects with AI tooling and prompts
- frogstar - Weight lifting planner/tracker (nextjs)
- vogon - Plex/Jellyfin art management (nextjs)
- hactar - A Plex hard drive savior (node)
- react-svg-path - React SVG path component library
- path - A simple svg path generation utility (SVG)
- A 15-year Old Homeless Kid Invented Neural Networks
- The Only Skill You Need to Improve Your Codebase
- Run MLX LLMs 50% Faster on a Mac with DSpark
- Run your AI on Nvidia’s GPUs for FREE
- Turn Claude Cowork into a Local AI Agent
- Run LLMs 2X faster on a Mac with oMLX
- Build a local coding agent from scratch with Ollama and AI SDK
- Build a local RAG system with Ollama, libSQL, and an EPUB
- What We Didn't Learn from Walter Pitts
- Baby's First Local AI
- Build a local RAG system with Ollama, libSQL, and an EPUB
- Ollama + Pi
- 2025 Movies You May Have Overlooked