Explore the directory

A web dashboard for managing several Ollama instances at once, letting a developer pull, update, filter and delete models across local and remote endpoints from one screen.

• Inference

An open-source mock server that generates a running API from an OpenAPI specification, validating both the spec and every incoming request against it, with optional latency and error simulation.

• Testing
• APIs

A simple wrapper that turns IndexedDB into a vector database directly in the browser, for storing and querying vector data without a server.

• Databases

A command-line tool that uses a tree-search approach guided by an LLM to iteratively rewrite code toward a measurable goal, such as optimizing an agent harness, a prompt, a training pipeline or a GPU kernel.

• Coding agents
• CLI tools

A streaming integrity monitor that detects decoding corruption in an LLM's output, such as repetition loops or language drift, while the answer is still generating, so the host can cut the stream and regenerate before the user sees it.

• AI evals

A private, Claude-Code-style coding agent for Apple Silicon that runs chat, code editing and local model workflows entirely on-device, compatible with Ollama and OpenAI APIs.

• Coding agents

An MCP server that lets Claude Code consult Google's Gemini model for complex coding problems, sending full code context and keeping the conversation going across follow-up questions.

• MCP servers

A collection of skills that teach AI coding assistants how to set up, test, deploy and debug LangChain, LangGraph, LangSmith and Deep Agents projects.

• Agent skills

A self-hosted retrieval-augmented-generation search engine that indexes 34 file formats with hybrid BM25 and semantic search, and can call Gemini, OpenAI, Claude or a local Ollama model to answer from the retrieved passages.

• RAG

A low-overhead LLM router written in Rust that fronts 80 providers and self-hosted vLLM or SGLang backends behind one OpenAI-compatible endpoint, reading existing LiteLLM-format configuration files unchanged.

• LLM gateways

A Chinese-language tutorial that builds a Text2SQL agent from scratch, covering state-graph orchestration, dual retrieval-augmented generation, human-in-the-loop confirmation and automatic SQL correction.

• Tutorials

A native macOS app that puts one question to several large language models at once, has them critique each other's answers anonymously, and shows where they agree or disagree before the user decides.

• Inference

A harness that makes AI coding tools such as Claude Code, Codex, Antigravity CLI and OpenCode work like a structured engineering team, with planning, testing, code review and security-audit skills, and one configuration across platforms.

• Coding agents
• Agent skills

A modified build of vLLM that adds Continuum-style request scheduling to improve inference throughput, meant as a base for tuning key-value cache pinning logic on different workloads.

• Inference

CODEC is the open-source engine behind Sovereign AI Workstation, a voice-controlled AI agent for macOS that watches the screen, controls apps, writes code, and works as both an MCP client and an MCP server.

• Agent frameworks
• MCP servers
• Speech

A Golang database manager and code generator with a command-line interface for SQL migrations, embedded SQL scripts, and generating models from an existing schema.

• CLI tools
• Databases

A macOS app that runs large language models such as DeepSeek V4 and Qwen3.8 on Apple Silicon by reading model experts from SSD, with a built-in chat interface and an OpenAI-compatible API.

• Inference

A self-hostable, open-source voice agent operating system with a Rust runtime tuned for low latency, an AI-driven agent builder, and a Python control plane, meant as an alternative to managed voice AI platforms.

• Speech

A Docker image that runs a self-hosted Whisper speech-to-text server with an OpenAI-compatible transcription and translation API, optional speaker diarization, and GPU acceleration.

• Speech
• Containers

A lightweight Python framework for building an AI agent in a few lines of code, with a tool decorator, streaming output, context management and native Model Context Protocol integration.

• Agent frameworks

A benchmark and toolkit for skill-retrieval-augmented LLM agents that treats reusable agent skills as a large corpus to retrieve from, rather than listing every skill in the prompt.

• RAG
• AI evals

An optimized SGLang-derived inference runtime tuned for running Qwen3.8 models on a single NVIDIA RTX PRO 6000 Blackwell GPU, with prebuilt launch configurations, a long context window, and disk-backed caching.

• Inference

Research code for a reinforcement-learning method that trains reasoning models with dense token-level supervision by contrasting teacher-student gaps under correct and incorrect reference solutions.

• Models
• Papers

A desktop app that runs many AI coding agents such as Claude, Codex, OpenCode, Gemini and Cursor side by side in one window, in terminal or chat layouts, with a built-in MCP server so agents can orchestrate each other and the app.

• Coding agents