3x faster speeds on MLX | Qwen 3.8 27B | Native MTP Speculative Decoding On Apple Silicon With No External Drafter.

Up to 3.4× faster LLM decoding on Apple Silicon, lossless. Native MLX port of DeepSeek's DSpark & z-lab's DFlash speculative decoding — Gemma-4, Qwen3.8, Muse-Glimmer, Nemotron, Ornith-1.0, ternary…

The fastest local AI engine for Apple Silicon. 4.2x faster than Ollama, 0.08s cached TTFT, 100% tool calling. 17 tool parsers, prompt cache, reasoning separation, cloud routing. Drop-in OpenAI repl…

Smart, Fast & Lossless session compaction for Pi. No LLM calls - produces structured, transcript-preserving summaries using pure extraction and formatting.

The @skills protocol: use any SKILL.md by path, save it into .atskills/, auto-trigger it with one line. Reference implementation + integration guide for any agent.

Package any component into a self-contained, isolated widget

TypeScript-to-Native Compiler

Headless control for Apple's Simulators — 3D models, taps, swipes, multi-finger gestures, 60 fps streaming, and a multi-device farm

Open-source native macOS HTTP debugging proxy — intercept HTTPS, inspect APIs, mock responses, debug WebSocket & GraphQL. Community-driven. For developers, by developers.

Marketing Skills, plus Mike's Personal Software Factory. More coming soon.

Run Service Workers anywhere

Build terminal apps with HTML, CSS and the DOM.

LoRA cold-start thinking SFT experiments on Nanbeige4-3B-Base using reasoning traces distilled from frontier models

Replacing lawyers with markdown files

A library of creative canvas components. Real HTML with WebGL effects running over it. React, Vue, Svelte, vanilla.

Own your AI. The native macOS harness for AI agents -- any model, persistent memory, autonomous execution, cryptographic identity. Built in Swift. Fully offline. Open source.

Rust GUI components for building fantastic cross-platform desktop application by using GPUI.

Fine-tune LLMs from one YAML. Layer streaming trains an 8B model on a 4 GB laptop GPU.

MiniMax H3 inference engine for Mac computers

LLM Council works together to answer your hardest questions

Open-source, self-hosted AI vulnerability research tool that orchestrates agents to find and validate security issues in code.

DiffuCoder: Understanding and Improving Masked Diffusion Models for Code Generation

Agentic QE Fleet is an open-source AI-powered QA/QE platform designed for use with Coding Agents (works best with Claude Code) featuring specialized agents and skills to support testing activities …

🌊 The original agent meta-harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence…

Agentic Chat App is an advanced AI-powered chat application designed for seamless real-time communication and intelligent responses. Built with Next.js, OpenAI, and Exa API, it leverages cutting-ed…

Open source agent with local models built in. Fully private and offline. Works out of the box on any hardware.

DoltLite - Version Controlled SQLite

Graph-Native Infrastructure for Context and Accountable AI Systems