Starred repositories
Guardrail capabilities for Pydantic AI — cost tracking, prompt injection detection, PII filtering, secret redaction, tool permissions, and async guardrails. Built on pydantic-ai's native capabiliti…
Open-source, self-hosted Claude Code - a terminal AI assistant and the Python framework behind it. Tool-calling, sandboxed execution, multi-agent teams, skills, checkpoints, unlimited context - on …
Full-stack AI app generator — FastAPI + Next.js with AI Agents, RAG, streaming, auth, and 20+ integrations out of the box.
A framework for building, orchestrating and deploying AI agents and multi-agent workflows with support for Python and .NET.
A 2.78-trillion-parameter Kimi K3 running inference on a single CPU in 8.24 GB of RAM. Portable C99: no BLAS, no framework, no GPU.
An all-in-one, pure C++ inference engine for audio models, powered by ggml. Supports TTS, STT, VAD, voice conversion, music generation, and more, with highly optimized performance. No Python depend…
A fully local, open-source voice-to-text tool that acts as a system-wide AI dictation layer, converting speech into clean, formatted text.
Voice-to-text with push-to-talk for Wayland compositors
A no-fluff and highly practical masterclass that reignites engineering curiosity and helps SDE-2, SDE-3, and above become great at designing, implementing, and shipping scalable, fault-tolerant, an…
NVIDIA Linux open GPU kernel module source
Multi-agent deep research for Claude Code. Zero API keys. Paste one line, type /research.
Professional Claude Code skills marketplace featuring production-ready skills for enhanced development workflows.
General-purpose deep research skill for AI agents — subagent-driven, source-backed, cited reports inline. skills.sh-compliant.
A curated list of awesome skills for Cursor
Deploy autonomous AI agents as your digital twins across 10 social platforms
Up to 3× faster LLM decoding on Apple Silicon, lossless. Native MLX port of DeepSeek's DSpark & z-lab's DFlash speculative decoding — Gemma-4, Qwen3, Ornith-1.0, ternary Bonsai-27B.
Build a compiler to solve Anthropic's interview challenge.
LMCache: Supercharge Your LLM with the Fastest KV Cache Layer
Inferno aims to be a super lightweight, highly efficient Rust inference engine for running open weights models on Apple Silicon with Metal, targeting machines such as a MacBook Pro with 64 GB of un…
Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦
LiteRT, successor to TensorFlow Lite. is Google's On-device framework for high-performance ML & GenAI deployment on edge platforms, via efficient conversion, runtime, and optimization
High-efficiency floating-point neural network inference operators for mobile, server, and Web
A close-to-metal Python API for programming AMD Ryzen™ AI NPUs (AI Engines), built on an open-source MLIR-based compiler toolchain.