Packages

Automated, regularly-updated curated collection of arXiv papers on causality and machine learning, with Python scripts for fetching and organizing new papers.

Home server operating system for self-hosting with an app store of 300+ apps, file management, backup, and external storage support

Byte Latent Transformer — a novel byte-level LLM architecture from Meta that eliminates tokenization while matching tokenized model performance, using entropy-driven dynamic patching to allocate compute based on data complexity.

Curated, weekly-updated collection of 150+ pre-built AI agent skills for the OpenClaw platform, covering AI tools, productivity, marketing, frontend/backend development, databases, auth, DevOps, and web automation.

Runs AI agent capabilities on a $5 ESP32-S3 microcontroller in pure C — no Linux, no Node.js. Provides a Telegram-based AI assistant with persistent memory, tool calling (web search, scheduling), and dual LLM provider support (Claude/OpenAI).

Kubernetes red team adversary emulation tool with 36 kubectl-based commands mapped to 13 MITRE ATT&CK tactics. Supports passive reconnaissance and active exploitation modes for assessing cluster security posture.

Blender add-on providing comprehensive tools for creating and editing Bezier curves, including interactive drawing, primitive shapes, snapping, boolean operations, and mesh conversion.

Open-source framework for enabling dynamic voice conversations with NPCs in existing games. Uses LLMs for contextual dialogue, facial recognition for NPC identification, emotion detection from webcam, and SadTalker for lip-synced facial animations overlaid on game footage. Works without game source code modification.

Python bot that downloads trending TikTok videos using an unofficial API and compiles them into a single video using FFmpeg. Automates the process of aggregating trending content.

A comprehensive guide and tutorial repository for mastering Claude Code, covering setup, prompt engineering, commands, hooks, workflows, automation, integrations, MCP servers, and the BMAD method.

Research project enabling Visual Language Model agents to interact with real computer screens through mouse and keyboard operations via VNC, implementing planning-execution-reflection loops for autonomous task completion.

An AI assistant framework built on LLMs that supports autonomous task planning, skill creation, long-term memory, and deployment across WeChat, Feishu, DingTalk, and web interfaces.

Curated discovery and one-click installation platform for Model Context Protocol (MCP) servers. Developers submit MCP servers via GitHub issues, and Cline users can browse, search, and install them directly from the marketplace.

Generates realistic talking face video sequences from speech audio or video input using a Disentangled Audio-Visual System (DAVS). An academic research project from NJU/CUHK demonstrating speech-driven facial animation.

Open-source enterprise AI knowledge base and MCP/A2A management platform with admin UI, user management, SSO, supporting 20+ LLM providers including ChatGPT, Claude, Llama, and Ollama.

SDK for Google DeepMind's Gemini Robotics models providing full lifecycle tooling including checkpoint access, model serving, evaluation, data uploading, finetuning, and an agent framework leveraging the Gemini Live API for physical robot control.

Open-source AI chatbot app powered by MCP, built with Next.js and Vercel AI SDK. Connects to MCP servers via HTTP/SSE transport, supports multiple AI providers (OpenAI, Anthropic, xAI), and provides streaming responses with tool integration.

A sample reference architecture demonstrating event-driven serverless microservices on AWS for an e-commerce backend, using Lambda, DynamoDB, EventBridge, AppSync, and Cognito. Explicitly not production-ready.

A curated list of resources, tools, skills, tutorials, and articles for OpenClaw — an open-source self-hosted AI agent supporting 50+ integrations across messaging platforms like WhatsApp, Telegram, Discord, and Slack.

An educational course curriculum for Cloud Applied Generative AI Engineering (GenEng). Provides structured learning materials covering foundational GenAI concepts, OpenAI APIs, the OPL Stack (OpenAI, Pinecone, LangChain), and practical application development.