Explore the directory

A React web app that uses the Google Generative AI (Gemini) API to turn a plain-language app idea into HTML and Tailwind CSS code for quick prototyping.

• Design tools

An experimental Android app for testing Google's on-device Gemini Nano model through the ML Kit Prompt API, built with Kotlin and Jetpack Compose.

• Inference
• Mobile

A set of Databricks scripts that build a corpus from GitHub repository content, train and serve a semantic search model on CPU, and summarize repository code with the Azure OpenAI API.

• RAG
• Data engineering

An open-source semantic layer and MCP server that sits between the AI tools a team uses and its data warehouse, answering questions in plain language with the SQL that produced the answer, a confidence grade, and a receipt.

• MCP servers
• Data engineering

A basic sample SAP Business Technology Platform application showing a semantic search engine built with the CAP model, SAP HANA Cloud's Vector Engine, and SAP's Generative AI Hub, with a SAPUI5 frontend.

• RAG
• Tutorials

A verification layer for AI coding agents working on Apple projects: it runs real Xcode builds and tests against the Swift an agent wrote, reconciles static analysis findings against that evidence, and returns a signed proof with the exact repairs still needed.

• Code quality
• Testing

A curated list of autonomous agent frameworks and products, alongside a small PyPI package that wraps AI agents in a common interface for asynchronous running, debugging, and deployment.

• Agent frameworks
• Curated lists

A local-first framework for building a persistent AI companion, or 'digital lifeform,' whose personality, memory, and permitted actions are stored as auditable, structured files rather than scattered prompts and caches.

• Agent frameworks
• Agent memory

A pure-Rust inference engine that runs GGUF language models directly on the GPU through WebGPU compute shaders, working across NVIDIA, AMD, Intel, integrated, and Apple Silicon GPUs without a C++ toolchain or llama.cpp.

• Inference

A self-hosted AI appliance for NVIDIA GPU boxes that installs an orchestrator per machine, serves models through OpenAI- and Anthropic-compatible APIs, and automatically clusters multiple nodes for cross-node tensor-parallel inference and fine-tuning.

• Inference

A self-hosted AI platform that puts inference, tool use, browser automation, image generation, speech synthesis, and other services behind one OpenAI-compatible endpoint, deployed with a single docker-compose up.

• MCP servers
• LLM gateways

An open specification for a machine-readable file, agents.txt, that lets a website declare which agent-interaction protocols it supports -- authorization, payments, MCP servers, skill packages, and A2A -- without embedding protocol implementation details.

• MCP servers
• APIs

A no-root Android build toolchain for Termux that compiles, signs, installs and launches real APK files entirely on an Android phone, bundling a native aarch64 aapt2 and a CLI/TUI that wraps Gradle, adb and multi-JDK management.

• CLI tools
• Mobile

An Elixir mix task that generates Elixir code from Protocol Buffers definitions by calling protoc directly, running in the local project's context instead of as a global protoc-gen-elixir plugin.

• Languages

A 128-bit universal hash function for fast file and data-stream integrity checks with cryptographic strength, formally proven in Lean to have a low collision rate and reaching up to 130 GB/s on modern desktop CPUs.

• Libraries

A fast, idiomatic Go client library for the Telegram MTProto API, covering bots and user accounts, with generated protocol types, a handler and middleware system, a plugin lifecycle, pluggable storage backends and MTProxy support.

• Libraries
• APIs

A 27B coding and tool-use language model family for agentic repository work, released both as a full-precision PEFT adapter on a Qwen3.6 base and as a standalone, packed-weight GGUF edition for native local inference.

• Models
• Inference

An open-source iOS Simulator replacement that restores the classic one-window-per-device Simulator workflow after Xcode 27 replaced it with Device Hub, adding deep links, push notifications, proxy certificates and a scripting CLI.

• Mobile

A deployment guide and benchmark for running the Qwen3.6-35B-A3B mixture-of-experts model with multi-token-prediction speculative decoding on a single NVIDIA L4 24GB GPU using the official llama.cpp Docker image.

• Inference
• Tutorials

A hardened Docker-in-Docker sandboxing environment, using gVisor kernel virtualization and a zero-trust Layer 7 proxy, for running untrusted coding agents such as Claude Code, Codex, Gemini CLI and OpenCode safely.

• Security
• Containers

An embeddable vector database written in Rust, for storing vectors alongside metadata, searching for similar records, and keeping the data inside your own application, with optional Python and WebAssembly interfaces.

• Databases

A locality-first document and vector database that lets applications place related data into explicit ring coordinates and retrieve the bounded nearby working set before ranking, filtering, or building LLM context, reducing the data scanned per query.

• Databases

A file-based toolkit that keeps a project's coding lessons and rules inside the repository, compiles them into a deterministic, offline lint rule set, and exposes a queryable index of past decisions for any AI coding agent.

• Agent skills
• Code quality

A personal AI assistant that runs as a single always-on Swift daemon on your own machine, pairing a private Telegram bot with a chosen LLM to remember facts, run scheduled and proactive tasks, and execute approved tools.

• Agent frameworks
• Workflow automation