Kyle McCleary

masquerade. HTML

DeepSpec. DeepSpec: a full-stack codebase for training and evaluating speculative decoding algorithms

arcgentica. An ARC-AGI solution using Agentica from Symbolica

awesome-tuis. List of projects that provide terminal user interfaces

oh-my-pi. ⌥ AI Coding agent for the terminal — hash-anchored edits, optimized tool harness, LSP, Python, browser, subagents, and more

i2cssh. csshX like ssh tool for iTerm2

ProRL-Agent-Server. Agentic RL on Any Harness at Scale

digial-ising. An all-digital, ring-oscillator-based Ising solver.

ThunderKittens. Tile primitives for speedy kernels

data-recipes.

NextFaster. A highly performant e-commerce template using Next.js

awesome-design-md. A collection of DESIGN.md files analysis by popular brand design systems. Drop one into your project and let coding agents generate a matching UI.

wterm. A terminal emulator for the web

hyperframes. Write HTML. Render video. Built for agents.

winghostty. Ghostty, now on Windows! 👻

gpu-tlb. C++

rtk. CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies

claw-code. An agent-managed museum exhibit, built in Rust with Gajae-Code / LazyCodex — developed and maintained with no human intervention.

t1code. T3Code, but in your terminal

parameter-golf. Train the smallest LM you can that fits in 16MB. Best model wins!

PaCoRe. PaCoRe: Learning to Scale Test-Time Compute with Parallel Coordinated Reasoning

RSA. Python

ghostty. 👻 Ghostty is a fast, feature-rich, and cross-platform terminal emulator that uses platform-native UI and GPU acceleration.

cmux. Open source Ghostty-based macOS terminal with vertical tabs and notifications for AI coding agents. Built for multitasking, organization, and programmability.

helios. TypeScript

Pixel-Perfect. Pixel Perfect UI is a precision-crafted UI library built for developers and designers who care about details.

hermes-agent-self-evolution. ⚒ Evolutionary self-improvement for Hermes Agent — optimize skills, prompts, and code using DSPy + GEPA

Robyn. Robyn is a Super Fast Async Python Web Framework with a Rust runtime.

t3code. TypeScript

hive. Multi-Agent Harness for Production AI

desloppify. Agent harness to make your slop code well-engineered and beautiful.

ssd. A lightweight inference engine supporting speculative speculative decoding (SSD).

drift_models. Python

prime-rl. Agentic RL Training at Scale

ludic. Ludic – an LLM-RL library for the era of experience

skydiscover. AI-Driven Scientific and Algorithmic Discovery

zoekt. Fast trigram based code search

armada. Kubernetes-native framework for bot and scraper orchestration

asupersync. Async runtime for Rust where correctness is structural: region-owned tasks, cancel-correct protocols, capability-gated effects, and deterministic replay testing

python-jose. A JOSE implementation in Python

TALOS. hardware accelerator for deep convolutional neural networks

canonical-multimodal-rep. Python

MeshTNC. MeshTNC is a tool for turning consumer grade LoRa radios into KISS TNC compatible packet radio modems

openclaw. Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞

Monolith. RLM for coding agent

mango. Practical and Powerful wayland compositor (dwm but wayland)

GOAD. game of active directory

Megatron-LM. Ongoing research training transformer models at scale

Megatron-LM. Ongoing research training transformer models at scale

Minimal-Drifting-Models. A minimal implementation of Drifting Models for 2D toy data. Unlike diffusion/flow models that iterate at inference, drifting models evolve the pushforward distribution during training and generate in a single forward pass (1-NFE). The drifting field V governs sample movement: V -> 0 as generated matches data.

CSC7809_FoundationModels. A repository of code examples to accompany the LSU CSC7809/7700/47000 course on AI foundation models.

beautiful-mermaid. TypeScript

Primus. A flexible and high-performance training framework designed for large-scale foundation model training on AMD GPUs

pi. AI agent toolkit: unified LLM API, agent loop, TUI, coding agent CLI

awesome-ai-residency. List of AI Residency Programs

codex-ralph. Shell

Premium-Font-Collection. Some premium font collection of popular brands.

modded-nanogpt. NanoGPT (124M) in 90 seconds

ART. Agent Reinforcement Trainer: train multi-step agents for real-world tasks using GRPO. Give your agents on-the-job training. Reinforcement learning for Qwen3.6, GPT-OSS, Llama, and more!

better-context. A better way to get up to date context on libraries/technologies in your projects

whisperX. WhisperX: Automatic Speech Recognition with Word-level Timestamps (& Diarization)

erdosproblems. A community database for the problems on the erdosproblems.com site

tui-generative-art. TypeScript

cc-mirror. Create multiple isolated Claude Code variants with custom providers (Z.ai, MiniMax, OpenRouter, LiteLLM)

jj. A Git-compatible VCS that is both simple and powerful

video2ascii. WebGL-powered React Component for video to ASCII conversion

opentui. OpenTUI is a library for building terminal user interfaces (TUIs)

descript-audio-codec. State-of-the-art audio codec with 90x compression factor. Supports 44.1kHz, 24kHz, and 16kHz mono/stereo audio.

dacvae. DACVAE

chandra. OCR model that handles complex tables, forms, handwriting with full layout.

stream-mdx. A Fast and Fully Configurable MDX Stream Renderer

oh-my-openagent. omo/lazycodex: The coding agent for tokenmaxxers;the one and only agent harness for complex codebases. For your Codex, for your OpenCode

beads. Beads - A memory upgrade for your coding agent

slime. slime is an LLM post-training framework for RL Scaling.

snort3. Snort++

spider. Streamline on-policy/off-policy distillation workflows in a few lines of code

suricata. Suricata is a network Intrusion Detection System, Intrusion Prevention System and Network Security Monitoring engine developed by the OISF and the Suricata community.

cutile-python. cuTile is a programming model for writing parallel kernels for NVIDIA GPUs

tptt. 😊 TPTT: Transforming Pretrained Transformers into Titans

Z-Image. Python

disco_rl. Accompanying code for "Discovering State-of-the-art Reinforcement Algorithms" Nature publication

alphafold3. AlphaFold 3 inference pipeline.

alphaevolve_repository_of_problems. Jupyter Notebook

ddex-suite. High-performance DDEX XML builder and parser with native bindings for TypeScript/JavaScript and Python. Built on a single Rust core for consistent behavior across all platforms.

codebuff. Generate code from the terminal!

MolecularNodes. Toolbox for molecular animations in Blender, powered by Geometry Nodes.

frp. A fast reverse proxy to help you expose a local server behind a NAT or firewall to the internet.

oracle. Ask the oracle when you're stuck. Invoke GPT-5 Pro with a custom context and files.

breach-torrents. Torrent files for data breaches used in research

jaxley. Differentiable neuron simulations with biophysical detail on CPU, GPU, or TPU.

mirage. Mirage Persistent Kernel: Compiling LLMs into a MegaKernel

rtl_433. Program to decode radio transmissions from devices on the ISM bands (and other frequencies)

ShinkaEvolve. ShinkaEvolve: Towards Open-Ended and Sample-Efficient Program Evolution 🧬

trl. Train transformer language models with reinforcement learning.

cursor-talk-to-figma-mcp. TalkToFigma: MCP integration between AI Agent (Cursor, Claude Code, Codex) and Figma, allowing Agentic AI to communicate with Figma for reading designs and modifying them programmatically.

figma-mcp-write-server. A Model Context Protocol (MCP) server for writing to Figma files

claudette. Claudette is Claude's friend

12-factor-agents. What are the principles we can use to build LLM-powered software that is actually good enough to put in the hands of production customers?

verl. verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework

gepa. Optimize prompts, code, and more with AI-powered Reflective Text Evolution

SnitchBench. TypeScript

uv. An extremely fast Python package and project manager, written in Rust.

mitmproxy. An interactive TLS-capable intercepting HTTP proxy for penetration testers and software developers.

effect. Build production-ready applications in TypeScript

crush. Glamourous agentic coding for all 💘

claude-code-reverse. A Tool to Visualize Claude Code's LLM Interactions

opencode. The open source coding agent.

winutil. Chris Titus Tech's Windows Utility - Install Programs, Tweaks, Fixes, and Updates

augment-swebench-agent. The #1 open-source SWE-bench Verified implementation

mixture_of_recursions. Mixture-of-Recursions: Learning Dynamic Recursive Depths for Adaptive Token-Level Computation (NeurIPS 2025)

endlessh. SSH tarpit that slowly sends an endless banner

HRM. Hierarchical Reasoning Model Official Release

firecracker. Secure and fast microVMs for serverless computing.

firecracker-containerd. firecracker-containerd enables containerd to manage containers as Firecracker microVMs

firecrawl. The API to search, scrape, and interact with the web at scale. 🔥

opencode. A powerful AI coding agent. Built for the terminal.

nautilus_trader. Production-grade Rust-native trading engine with deterministic event-driven architecture

gemini-fullstack-langgraph-quickstart. Get started with building Fullstack Agents using Gemini 2.5 and LangGraph

exllamav3. An optimized quantization and inference library for running LLMs locally on modern consumer-class GPUs

HVM2. A massively parallel, optimal functional runtime in Rust

markdown-it. Markdown parser, done right. 100% CommonMark support, extensions, syntax plugins & high speed

WheelWizard. WheelWizard, Retro Rewind Launcher

moment-research. Research Code for MOMENT

adam-atan2-pytorch. Implementation of the proposed Adam-atan2 from Google Deepmind in Pytorch

servers. Model Context Protocol Servers

native-sparse-attention-pytorch. Implementation of the sparse attention pattern proposed by the Deepseek team in their "Native Sparse Attention" paper

titans-pytorch. Unofficial implementation of Titans, SOTA memory for transformers, in Pytorch

fashion-mnist. A MNIST-like fashion product database. Benchmark :point_down:

insanely-fast-whisper. Jupyter Notebook

pyespargos. The python library for the ESPARGOS WiFi antenna array

hypersubatomic-vscode-theme. Hypersubatomic is a high contrast UI and syntax theme for VSCode

slacktyping. i'm typing when you're typing

Balatro-ModpackManager. Manage Balatro mods effortlessly with the Modpack Manager. Install, update, and customize modpacks in one streamlined tool.

sunnypilot. sunnypilot is an open source driver assistance system. sunnypilot offers the user a unique driving experience for over 350 supported car makes and models with modified behaviors of driving assist engagements. sunnypilot complies with the safety policy from comma.ai's openpilot as accurately as possible.

bitspace. A prototype of a creative environment for the 21st century

monolith. A Lightweight Recommendation System

deno. A modern runtime for JavaScript and TypeScript.

genesis-world. Simulation platform for general-purpose robotics & embodied AI learning.

bonsai. A library for building dynamic webapps, using Js_of_ocaml

TRELLIS. Official repo for paper "Structured 3D Latents for Scalable and Versatile 3D Generation" (CVPR'25 Spotlight).

mozilla-type-family. A collection of custom fonts for Mozilla

dive. A tool for exploring each layer in a docker image

openv0. AI generated UI components

optimizers. For optimization algorithm research and development.

causal-conv1d. Causal depthwise conv1d in CUDA, with a PyTorch interface

teable. ✨ The Next Gen Airtable Alternative: No-Code Postgres

StyleTTS2. StyleTTS 2: Towards Human-Level Text-to-Speech through Style Diffusion and Adversarial Training with Large Speech Language Models

mdx. Markdown for the component era

docling. Get your documents ready for gen AI

sylph. A minimal portfolio template.

manim. A community-maintained Python framework for creating mathematical animations.

Awesome-LLM-Strawberry. A collection of LLM papers, blogs, and projects, with a focus on OpenAI o1 🍓 and reasoning techniques.

TikZ. Complete collection of my PGF/TikZ figures.

number-flow. An animated number component for React, Vue, Svelte, and TS/JS.

tabled. Detect and extract tables to markdown and csv

ImHex. 🔍 A Hex Editor for Reverse Engineers, Programmers and people who value their retinas when working at 3 AM.

gvisor. Application Kernel for Containers

open-parse. Improved file parsing for LLM’s

GOT-OCR2.0. Official code implementation of General OCR Theory: Towards OCR-2.0 via a Unified End-to-end Model

granite-tsfm. Foundation Models for Time Series

Conference-Acceptance-Rate. Acceptance rates for the major AI conferences

baml. The programming language for agents

plugin.video.seren. Repository for Seren Development

sglang. SGLang is a high-performance serving framework for large language models and multimodal models.

SqueezeLLM. [ICML 2024] SqueezeLLM: Dense-and-Sparse Quantization

langflow. Langflow is a powerful tool for building and deploying AI-powered agents and workflows.

ml-mkqa. We introduce MKQA, an open-domain question answering evaluation set comprising 10k question-answer pairs aligned across 26 typologically diverse languages (260k question-answer pairs in total). The goal of this dataset is to provide a challenging benchmark for question answering quality across a wide set of languages. Please refer to our paper for details, MKQA: A Linguistically Diverse Benchmark for Multilingual Open Domain Question Answering

KALMV. Official Code Repository for Knowledge-Augmented Language Model Verification (EMNLP 2023)

FLARE. Forward-Looking Active REtrieval-augmented generation (FLARE)

abliterator. Simple Python library/structure to ablate features in LLMs which are supported by TransformerLens

moment. MOMENT: A Family of Open Time-series Foundation Models, ICML'24

tabbyAPI. The official API server for Exllama. OAI compatible, lightweight, and fast.

sonar. Large-scale LLM inference engine

pgvectorscale. Postgres extension for vector search (DiskANN), complements pgvector for performance and scale. Postgres OSS licensed.

MultiHop-RAG. Repository for "MultiHop-RAG: A Dataset for Evaluating Retrieval-Augmented Generation Across Documents" (COLM 2024)

2wikimultihop. Python

UXM. Enables file modding for DS2, SotFS, DS3, and Sekiro by unpacking game archives and patching the executable to load loose files instead.

pyppeteer. Headless chrome/chromium automation library (unofficial port of puppeteer)

G-Retriever. Official Implementation of NeurIPS 2024 paper "G-Retriever: Retrieval-Augmented Generation for Textual Graph Understanding and Question Answering""

unsloth. Unsloth Studio is a web UI for training and running open models like Gemma 4, Qwen3.6, DeepSeek, gpt-oss locally.

CodeR.

mamba.np. A pure NumPy implementation of Mamba.

geoalg. Geometric Algebra Calculators - quaternions, octonions, sedenions and Clifford algebra

simplie. a simple program for Lie algebras

hamilton. Apache Hamilton helps data scientists and engineers define testable, modular, self-documenting dataflows, that encode lineage/tracing and metadata. Runs and scales everywhere python does.

monaco-react. Monaco Editor for React - use the monaco-editor in any React application without needing to use webpack (or rollup/parcel/etc) configuration files / plugins

MoRA. MoRA: High-Rank Updating for Parameter-Efficient Fine-Tuning

bleurt-pytorch. BLEURT implementation in PyTorch

drizzle-orm. ORM

unofficial-apis. A collection of unofficial apis. Designed to inspire your next Friday night hack

arc90-readability. A copy of the original Arc90 repo with links to many of the current ports.

python-readability. fast python port of arc90's readability tool, updated to match latest readability.js!

next-themes. Perfect Next.js dark mode in 2 lines of code. Support System preference and any other theme with no flashing

instructor. structured outputs for llms

llama3. The official Meta Llama 3 GitHub site

ComfyUI-Nodes-Visualization. 2400+节点可视化 Visualization | Collection of ComfyUI Custom Nodes

orchestrator. MySQL replication topology management and HA

dataset-distillation. Soft-Label Dataset Distillation and Text Dataset Distillation

LO-Shot. Papers and code related to 'Less Than One'-Shot (LO-Shot) Learning

ELLA. ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment