Explore the directory

A lightweight Ruby framework for building AI agent workflows as a state machine, with a pluggable agent and I/O interface and an OpenAI-backed agent implementation included.

• Agent frameworks

A WinUI 3 sample chat app from Microsoft that runs Windows' on-device Aion Instruct Preview language model on Copilot+ PCs, streaming tokens with first-token latency and tokens-per-second shown under each reply.

• Inference
• Tutorials

A self-evolving AI security research platform that publishes code-fix tasks through a honeypot, evaluates submitted fixes in a Docker sandbox with cheat detection, and exports the results as training data for LoRA fine-tuning.

• AI evals
• Security

An open-source, self-hostable Next.js SaaS boilerplate for building custom AI chatbots trained on your own documents, URLs, and Q&A data, with citations, embeddable widgets, and built-in billing.

• RAG
• Web frameworks

A demonstration enterprise-grade retrieval-augmented generation system with a Next.js chat frontend, a FastAPI backend, and a LangChain AI layer, built for on-premise deployment in data-sensitive industries such as banking and healthcare.

• RAG
• Tutorials

A typed Kotlin DSL runtime for building AI agent systems on the JVM, where the compiler enforces input/output types and tool permissions, and every run produces a tamper-evident audit trail.

• Agent frameworks

A free hook that speaks a short summary out loud after your coding agent finishes a reply, using a local ONNX text-to-speech model with no cloud API, API key, or subscription.

• Agent skills
• Speech

Generates an OpenAPI 3.1 specification for a Phoenix application automatically, by reading its Ecto schemas, JSON views and router instead of a separate spec file.

• Libraries
• APIs

A terminal UI for browsing, inspecting and pruning vector databases such as Qdrant, Pinecone and Chroma, without leaving the terminal.

• CLI tools
• Databases

An unofficial documentation site and knowledge base for developers building on JAM, the Polkadot ecosystem's next-generation protocol.

• Documentation tools

A .NET binding for Google DeepMind's EmbeddingGemma-300m model that runs fully local, offline text embedding with no API key or cloud dependency, distributed as a single NuGet package for ASP.NET Core and Microsoft Semantic Kernel.

• RAG
• Libraries

A curated list of tools for exploring the dark web: search engines, onion-link directories, onion scanners, dark-web crawlers and cyber threat intelligence resources, linking out to each individual tool's own repository.

• Security
• Curated lists

An example Apache Airflow project that orchestrates a daily ETL pipeline fetching weather data from a free API, cleaning and transforming it, and storing it in PostgreSQL, ready for downstream machine learning tasks.

• Data engineering
• Tutorials

A code generation engine that scaffolds production-ready React, Vue, Angular and React Native components styled with CSS Modules, SCSS, Tailwind, NativeWind or StyleSheet, producing plain source files a project owns instead of a runtime package.

• CLI tools
• UI kits

A full-stack GraphRAG application with a Django backend and React frontend that builds a knowledge graph from ingested documents, detects thematic communities, and answers questions using hybrid vector and graph retrieval with LLM-powered reasoning.

• RAG

A standalone desktop app for Editor Code Assistant (ECA), an AI pair-programming system, that connects to an external eca server for chat, code suggestions and context management without requiring an editor at all.

• Coding agents

A compact, exact-scan vector database distributed as a single executable, keeping vectors resident on an NVIDIA GPU (or system RAM in a CPU build) and reachable through SDKs for C++, Python, Node.js, TypeScript, and Delphi.

• Databases

A native desktop workspace that runs 27 external AI coding CLIs alongside a built-in agent, developer tool CLIs, an embedded editor, an in-app browser, and a layered image editor, replacing a scattered set of terminal tabs with one command center.

• Coding agents
• Terminals

An independent workspace and harness for AI coding and research agents, usable on desktop, web, and mobile, built around a bundled Pi runtime with a governed permission system and two switchable interfaces: a session-centered agent workspace and a full IDE workbench.

• Coding agents

A V language binding for llama.cpp, letting V programs run quantized large language models such as DeepSeek or Qwen locally on CPU hardware.

• Inference
• Libraries

A typed, auditable Rust database runtime and query builder built for AI coding agents, enforcing agent identity, declared intent, and redacted diagnostic logging around every database operation.

• Security
• Databases

An MCP server that connects China's electronic fapiao invoicing system to AI models such as Claude, GPT, Cursor, and Cherry Studio, letting them issue, query, void, and verify invoices through natural-language requests.

• MCP servers

A dual-mode, LangGraph-based security platform: a red-team agent that chains reconnaissance, exploitation, and reporting with terminal-verified exploits, and a blue-team agent that monitors live HTTP traffic and responds with deception, blocking, or patching.

• Security

A self-hosted, full-stack retrieval-augmented generation system with a FastAPI backend built on pydantic-ai, a Flutter client, and a terminal UI, deployable by having an AI coding agent install its published Agent Skills.

• RAG