A library of abstractions and recipes for fine-tuning language models via the Tinker API, supporting SFT, RL, and multimodal training.

A high-performance library for constrained decoding that ensures LLMs produce 100% structurally correct outputs like JSON and regex with near-zero overhead.

An MCP server that enables AI agents to interact with Grafana, allowing them to query datasources, manage dashboards, and handle observability operations.

A visual review tool that lets users edit HTML and Markdown files directly and send batch feedback to AI agents instead of using chat prompts.

A local proxy that bridges the Claude Agent SDK to the standard Anthropic and OpenAI APIs, allowing third-party AI agents to use Claude Max subscription access.

A high-performance AI gateway that provides a unified API for multiple LLM providers with built-in cost tracking, caching, and failover mechanisms.

An open-source framework by Google for building full-stack AI applications with unified APIs for multiple model providers and integrated developer tools for testing and monitoring.

A self-hosted inference server for Apple Silicon that provides OpenAI and Anthropic compatible APIs for running LLMs, VLMs, and image generation models locally.

A zero-dependency SVG and JavaScript expression engine that provides 32 emotional states for AI assistants, chatbots, and desktop pets.

A 15TB collection of diverse numerical physics simulations and a benchmarking suite designed to train and evaluate machine learning surrogate models for computational sciences.

An AI-powered tool for scanning and diagnosing Kubernetes cluster issues, providing plain-English explanations and fix recommendations.

A collection of Model Context Protocol (MCP) servers and core tooling that standardizes how AI agents and LLMs connect to Microsoft services and external data sources.

A unified proxy server for AI CLI tools that consolidates multiple AI subscriptions and API keys into a single managed endpoint with routing, quotas, and observability.

A document processing toolkit for Korean administrative files (HWP, HWPX, PDF) that converts them to Markdown for AI analysis and generates official government-standard documents.

A framework for efficient LLM inference and fine-tuning that uses CPU-GPU heterogeneous computing to run ultra-large MoE models on consumer hardware.

ExecuTorch is an open‑source PyTorch‑based runtime for deploying AI models on edge devices—from phones to micro‑controllers. It exports models directly from PyTorch, compiles them ahead‑of‑time into a tiny `.pte` file, and runs them on a variety of hardware back‑ends via lightweight C++/Python/Swift/Kotlin APIs. It supports vision, speech, and LLMs (e.g., Llama) and is used in Meta’s production apps.

An open-source vulnerability scanner for LLMs and agent workflows that protects AI systems from jailbreaks, fuzzing, and multimodal attacks.

A Docker CLI plugin that provides a gateway for managing and deploying isolated, containerized MCP servers for AI clients.

A framework that combines LLMs and evolutionary algorithms to automate the discovery and optimization of scientific code across various programming languages.

A system-level observability tool for AI agents that uses eBPF to monitor model calls, process execution, and file activity without requiring an SDK or proxy.

Agentic Coding Flywheel Setup (ACFS) is a Bash installer that turns a fresh Ubuntu/Arch VPS into a fully‑configured AI‑coding workstation. It installs a modern shell, language runtimes, cloud CLIs, and three AI coding agents (Claude, Codex, Antigravity) via a single, idempotent `curl | bash` command. The toolchain is defined in a YAML manifest, from which TypeScript code generates the installer scripts. A companion Next.js wizard guides beginners through VPS provisioning and SSH setup. Security is handled with HTTPS‑only downloads and SHA‑256 checksum verification. The project is MIT‑licensed and targets both individual developers and teams needing reproducible agentic environments.

An SQLite extension that adds graph database capabilities and Cypher query language support to SQLite, enabling embedded graph data storage and querying.

A management layer for AI coding agents that provides isolated workspaces via Git worktrees and a unified terminal interface for any CLI-based agent.

A Python framework for building production-ready conversational AI applications with a built-in UI, reducing development time from weeks to minutes.

A one-line Exploratory Data Analysis (EDA) tool that generates comprehensive profiling reports for pandas and Spark DataFrames, identifying data quality issues and patterns.

A proposal to standardize the use of /llms.txt files to provide AI agents with concise, Markdown-formatted summaries and links to clean content on websites.

A configuration and component manager for OpenCode that enables users to apply portable, SHA-verified profiles and components across different repositories.

A lightweight, rootless Linux process sandbox that uses Landlock and seccomp to confine untrusted code with minimal overhead and fast startup.

A proxy that enables MCP clients supporting only local stdio servers to connect to remote MCP servers with full OAuth authorization support.

An on-device Gen AI inference runtime for Qualcomm Snapdragon devices that allows users to run LLMs and VLMs locally using GGUF models or optimized Qualcomm AI Hub bundles.

A CLI tool that enables different AI coding agents to message, observe, and spawn each other across terminals to coordinate complex development pipelines.

A browser automation tool that allows AI coding agents to maintain persistent browser sessions and use compact accessibility trees for efficient web interaction.

An AI security platform combining a three-layer code auditing system (Shield) and an automated, authorized penetration testing engine (Spear).

A Godot Engine plugin that implements the Model Context Protocol (MCP), enabling AI assistants to read and modify scenes, scripts, and resources directly via natural language.

An MCP server that allows AI agents to automate Firefox using WebDriver BiDi, enabling them to navigate the web, interact with elements, and access developer tools.

A JavaScript knowledge graph querying framework that enables SPARQL and GraphQL queries over decentralized RDF data on the Web.

A Dashboard-as-Code tool that lets users and AI agents define, validate, and serve dashboards using YAML or TSX with a built-in semantic layer.

An open-source framework for building AI-powered offensive and defensive security automation, used to establish the Cybersecurity AI research domain.

A knowledge base of skills, rules, and validation hooks that teach AI coding agents to generate correct, compile-ready UdonSharp code for VRChat world development.

A universal, serverless pasteboard that enables real-time, end-to-end encrypted clipboard sharing across Windows, macOS, and Linux devices over a local network.

mlx‑dspark is a Python package (and macOS app) that runs loss‑less speculative decoding (DSpark & DFlash) for many popular LLMs on Apple‑Silicon Macs via MLX, delivering 2‑4× faster generation and exposing an OpenAI‑compatible local API.

An interactive coding platform for practicing PyTorch implementations of operators and architectures from scratch to prepare for ML engineering interviews.

An inference engine and training environment for Recursive Language Models (RLMs) that enable LLMs to handle near-infinite context by programmatically decomposing tasks via recursive sub-calls in a REPL environment.

agent‑style is a Python/Node library that supplies 21 classic‑and‑LLM‑observed writing rules for AI coding‑and‑writing agents. The rules are injected into the model’s system prompt (soft enforcement) and can also be applied as a post‑generation review skill. Benchmarks on Claude Opus, GPT‑5.4, and Gemini 3 Flash show sizable drops in mechanical rule violations, proving the approach can make AI‑generated technical prose clearer.

A lightweight, high-performance text chunking library for RAG pipelines that offers multiple splitting strategies and extensive integrations with vector databases and embedding providers.

A human-centric framework for building and manage real-life AI and ML systems, streamlining the transition from prototyping to production deployment.

A browser automation skill for coding agents that enables them to write and execute custom Playwright scripts for testing and complex web flows.

An HTTP/HTTPS proxy that provides security guardrails for AI agents by evaluating outbound API requests against static rules and LLM-based policies.