Explore the directory

A portable Rust library for neural network training and inference that runs on Vulkan and Metal through blade-graphics instead of CUDA, compiling models to specialized WGSL shaders with automatic graph rewriting.

• Models
• Libraries

An ICML 2026 benchmark and a family of domain-specialized vision-language models for understanding mechanical engineering drawings, covering recognition, reasoning and judging tasks over dense technical diagrams.

• Models
• AI evals

An MCP server that connects AI agents to Decodo's web scraping API, letting tools such as Claude, Cursor and Windsurf scrape websites, search engines, e-commerce pages and social media without building scraping infrastructure.

• MCP servers
• Browser automation

Companion study material for a book preparing readers for the NVIDIA NCP-AAI agentic AI certification, with chapter text, quizzes, practice exams, slides, labs and mappings to several other AI certifications.

• Tutorials
• Books

An optimized local serving engine for the Kimi-Linear-48B model on a single consumer GPU, using INT4 quantization and fused decode kernels, with a protocol bridge that lets Claude Code, Codex, Cline, Aider and opencode connect to it unchanged.

• Inference

A lightweight RAG provenance middleware that checks, without calling an LLM, whether every claim in a generated answer is actually grounded in a retrieved source, and flags claims backed only by outdated sources.

• RAG
• AI evals

A step-by-step guide, with an automated installer script, for running the DeepSeek R1 model offline on a rooted Android phone using Termux, Arch Linux and Ollama.

• Inference
• Tutorials

A TypeScript framework for building tool-using AI agents, from a single scripted agent to a multi-agent swarm, with a scaffolding CLI, an interactive setup wizard, and support for more than ten LLM providers.

• Agent frameworks

A community-curated list of frameworks, tools, libraries and platforms for building and deploying AI agents, covering coding, research, creative, web and voice agents plus memory, evaluation and deployment tooling.

• Agent frameworks
• Curated lists

An auto-updated curated list of AI agent skill repositories for Claude, GPT and other agents, refreshed every 15 minutes by a GitHub Actions crawler that tests and rates each one by quality and maintenance status.

• Agent skills
• Curated lists

A cross-platform desktop traffic-light widget that shows the live status of Claude Code and Codex sessions grouped by project, turning red when input is needed, yellow while the agent works and green when it finishes.

• Coding agents
• Desktop utilities

Course material and a working proof-of-concept for an O'Reilly live training that prepares learners for Microsoft's AB-100 agentic AI certification exam, plus a Copilot agent that generates practice questions.

• Tutorials

A beginner-friendly Python toolbox of simple, plain-English-style functions for common tasks such as file handling, text processing, numbers, dates and basic web and image operations, meant to keep syntax out of the way for new programmers.

• Libraries

A zero-dependency Java library of sorted Set and Map implementations, including AVL trees, red-black trees, B-trees and B+ trees, built for cache locality and compliant with JDK 21's SequencedCollection interfaces.

• Libraries

A zero-token, long-term memory extension for the pi coding agent that reimplements a published Zero-Mem design, remembering prior sessions without spending extra LLM calls on retrieval.

• Agent memory

A VS Code extension for a transparent AI coding assistant, built on the Pi engine, that traces generated code back to the agent session and reasoning that produced it and shows every tool call.

• Coding agents

A tool that turns Claude, Cursor, Codex and other AI coding sessions into a single shareable, animated HTML replay of every prompt, thought and tool call, with local analytics and search across all sessions.

• Coding agents
• Debugging

A Python library for creating, simulating and formally verifying scenarios with multiple interacting, decision-making agents, such as autonomous vehicles, using hybrid reachability analysis.

• Testing
• Libraries

A cryptographic trust layer for AI agent actions that signs every tool call as a hash-chained, offline-verifiable receipt, so an agent's actions and the authorization behind them can be proven rather than only logged.

• Agent frameworks
• Security

Reference implementations from Vector Institute's RAG bootcamp that show common retrieval-augmented generation methodologies over web, document, SQL and cloud data sources using LangChain and LlamaIndex.

• RAG
• Tutorials

A small Ruby AI agent framework with an async message bus, multiple LLM providers, CLI, Telegram and WhatsApp channels, scheduling, browser automation, and separate memory personas per workspace.

• Agent frameworks

A single-file PyTorch implementation of MoonViT, the native-resolution vision encoder from Kimi-VL, that packs images of different sizes into one sequence instead of resizing or padding them.

• Models
• Libraries

A packaging of a durable local inference server for the Qwen3.8 27B model on a single consumer NVIDIA GPU, built to resume a coding agent's session from saved state after the inference server restarts.

• Coding agents
• Inference

A benchmarking framework for single- and multi-agent LLM systems that pairs built-in datasets with evaluators, visual debugging of tool use and accuracy, and automated workflow optimization.

• Agent frameworks
• AI evals