Explore the directory

A hands-on workshop that builds a multi-agent predictive-maintenance workflow for a fictitious tire factory using Microsoft Agent Framework, Microsoft Foundry and Azure.

• Tutorials

A vLLM component that provides custom SYCL kernels for Intel GPUs, covering normalization, attention, quantization and mixture-of-experts operations, to accelerate large-language-model inference on Intel hardware.

• Inference

A desktop and mobile front end for tmux that adds a graphical terminal, file explorer, Git view and agent-status sidebar over an existing tmux server, so coding-agent sessions like Codex or Claude Code stay reachable from a phone.

• Coding agents
• Terminals

An open-source Python library that simulates quantum circuits locally, so quantum programs written for the Amazon Braket SDK can be tested on a laptop before being submitted to the Braket cloud service.

• Testing
• Libraries

An engineering-focused AI code generation platform built on Spring Boot, LangChain4j and LangGraph4j, with a visible task plan, layered context compression and one-click local deployment.

• Coding agents

A coding-agent CLI and strongly-typed Python library that supports async tool calls, hot-swapping the model or provider mid-session, and bidirectional recursive agent spawning and messaging.

• Coding agents
• Agent frameworks

A deterministic spec-orchestration tool for the pi coding agent that drives a request through fixed refine, research, grill, compose and critique phases before handing the main session a verified spec to execute.

• Coding agents
• Agent skills

Client SDKs for Synap, a hosted memory layer that gives AI agents long-term memory across conversations, sessions and devices, with native integrations for major agent frameworks.

• Agent memory

A git template for building a personal LLM-maintained wiki, implementing Andrej Karpathy's three-layer pattern of source documents, an LLM-maintained wiki and a schema config file.

• Agent memory
• Documentation tools

A native macOS app that puts an entire small-language-model training workflow, from architecture to tokenizer to fine-tuning to chat, in one place, using Apple Silicon and MLX.

• Models

A self-hosted home AI inference platform for AMD Strix Halo hardware that turns a Linux box into an OpenAI-compatible endpoint for chat, embeddings, reranking, transcription, speech and image generation.

• Inference

A post-training quantization method for large language models that learns per-coordinate grid assignments and per-group scales through a Gumbel-Softmax relaxation, targeting 1 to 3 bits per parameter.

• Inference

A research query engine that uses a pipeline of LLM agents to generate instance-optimized, standalone executable code for SQL queries, tailored to a specific dataset, workload and hardware.

• Databases

A desktop delegation environment for running and supervising several AI coding agents in parallel, each in its own git worktree, with an assistant that can spawn, broadcast to and monitor them.

• Coding agents

A fork of OpenAI's Codex CLI that adds multi-agent orchestration, letting a parent agent spawn specialized child agents that work in parallel and report progress back.

• Coding agents
• Agent frameworks

A local-first CLI that hands off session context between AI coding agents, converting a saved session into Markdown so work can resume in the same agent or a different one.

• Coding agents

A fully offline chatbot that runs entirely on an ESP32-S3 Cardputer microcontroller, with an 8-million-parameter model quantized to 4 bits and baked into the firmware.

• Inference

An interactive canvas for inspecting a large language model's weights tensor by tensor, showing measured INT8, INT4 and FP8 quantization error, spectra and outlier channels for every tensor in a checkpoint.

• Models

A ComfyUI extension that adds workflow and node-favorite management, theme customization, alignment and grouping tools, plus image, video, audio and large-model nodes including offline text-to-speech and Qwen-VL inference.

• Generative media

A course that teaches vector databases, embeddings and retrieval-augmented generation from fundamentals to production, working through ChromaDB, Pinecone and LangChain in runnable Python scripts.

• RAG
• Tutorials

A hands-on course that builds a retrieval-augmented-generation system from scratch in plain Python, one Jupyter notebook at a time, before showing the same pipeline rebuilt with LangChain.

• RAG
• Tutorials

A modular Python client-server implementation of the Model Context Protocol, demonstrating an LLM client bridged to a weather API server and a document RAG server over stdio.

• MCP servers

A semantic code search tool for coding agents that scores code fragments with an AI judgment layer instead of exact-string matching, returning source excerpts with file paths and line numbers.

• Agent memory

A local-first, browser-based console for generating and editing images through any OpenAI-compatible image API, with request queuing, multi-endpoint debugging and prompt history.

• Generative media