from vibe coding to agentic engineering - practice makes claude perfect

vLLM Metal plugin powered by mlx-swift — high-performance LLM inference on Apple Silicon

35 domain-expert LoRAs on Qwen3.6-35B-A3B (MoE, 256 experts, 3B active). Cognitive layer: Aeon memory, CAMP negotiator, KnowBias. MLX on Mac Studio, Q4_K_M inference. Apache-2.0.

WhatsApp in pure Rust. Single binary, 5 MB, 15 MB RAM. 54 API endpoints, 30 MCP tools, full protocol.

AI agent toolkit: coding agent CLI, unified LLM API, TUI & web UI libraries, Slack bot, vLLM pods

Orchestrate an entire AI dev team on as little as 5GB VRAM. An AI coding agent built like a systems engineer. Ephemeral context, zero token bloat, exact-match diffs. Stop wasting money on 10k token…

Open-source platform for creating, distributing and running sovereign EU-compliant LLMs. Verticalize any model for your domain, language and brand. AI Act ready.

Tree-based speculative decoding for Apple Silicon (MLX). ~10-15% faster than DFlash on code, ~1.5x over autoregressive. First MLX port with custom Metal kernels for hybrid model support.

Lossless DFlash speculative decoding for MLX on Apple Silicon

MCP server for interacting with the iOS simulator

LLMs and VLMs with MLX Swift

Apple Silicon (MLX) port of Karpathy's autoresearch — autonomous AI research loops on Mac, no PyTorch required.

Native macOS App Store Connect tool with MCP. Submit iOS apps to App Store with AI agents

AirDrop-like file sharing for any device, anywhere.

Ultra-Sparse Adaptation of 1-Bit LLMs via XOR Patches

Drop-in replacement for Apache Spark UI

VS Code rebuilt on Tauri. Same architecture, 96% smaller. Early release.

Awesome Software Made in Germany

Docker-compatible container CLI built on Apple's Containerization framework. Same commands, same flags — mocker run, ps, stop, build, compose, stats — all working on macOS 26.

vMLX - Home of JANG_Q - Cont Batch, Prefix, Paged, KV Cache Quant, VL - Powers MLX Studio. Image gen/edit, OpenAI/Anth

Powdered Metal — High performance LLM fine-tuning framework for Apple Silicon

Apple Neural Engine (ANE) LLM inference engine — reverse-engineered private APIs, Metal GPU shaders, hybrid ANE+GPU+CPU on Apple Silicon. 32 tok/s matching llama.cpp, 3.6 TFLOPS fused ANE mega-kern…

Train and run transformers directly on Apple's Neural Engine in Swift bypass coreml entirely

LLM inference server with continuous batching & SSD caching for Apple Silicon — managed from the macOS menu bar

HuggingFace transformer compiler for optimised native inference binaries

Multi-agent orchestration for Claude Code. Persistent memory, tasks, rules, and skills that make AI agents actually coordinate.

AI agents running research on single-GPU nanochat training automatically

Beautiful UIs in Rust. Cross-platform. Dead simple.