Explore the directory

A local, file-backed vector database for Node.js with an optional gRPC server for cross-language access, aimed at applications that want a Pinecone-like index stored on disk.

• Databases

mimo2codex is a local proxy that lets the latest OpenAI Codex CLI or desktop app talk to other large language models, with built-in support for Xiaomi MiMo and DeepSeek and a generic mechanism for any OpenAI-compatible or native Responses API provider.

• LLM gateways

StarNet is a local-first desktop application that runs real AI agents inside a pixel-art space station, where rooms represent capability-scoped teams and the layout you build is the actual multi-agent workflow that executes.

• Agent frameworks

Maestro is a local, one-click AI video, image and music studio built on the WanGP pipeline, with a director mode that plans a full music video or short film, screenplay, shots and clips, from a single prompt.

• Generative media

ChatLLM Web is a browser-based, private local model studio that runs chat and agent workflows entirely in the browser using WebGPU and WebLLM, with no backend, account or telemetry.

• Inference

AgentiLoop Agent! is a native macOS app that wires 23 language model providers, including on-device Apple Intelligence, into an autonomous task loop that can read a codebase, fix bugs, build an Xcode project, drive other Mac apps, and run shell commands.

• Coding agents

A free QR code generation service reachable entirely through curl, turning any text or file piped to it into a QR code as an image, SVG, or plain text, with configurable size, color, and error-correction level.

• CLI tools
• APIs

Casbin Gateway is a local dashboard and proxy that shows what every AI coding agent on a machine is doing, grades whether an API key's provider matches what was paid for, and centralizes model access, permissions and spend tracking across 44 vendors.

• LLM gateways
• Monitoring

A Python library that generates high-performance CPU and GPU code for array-based computations, such as dense linear algebra and PDE solvers, by transforming a simple description under the user's control.

• Libraries

SigMap builds a deterministic, line-anchored map of a codebase's signatures so AI coding agents and reviewers can retrieve relevant code and check that an AI's claims are grounded in real files and line numbers, without calling an LLM or embeddings.

• Agent memory

apktool-mcp-server is an MCP server built on Apktool that lets an LLM decompile Android APKs, inspect manifests and smali code, search and modify resources, and rebuild the APK, for reverse engineering and vulnerability analysis.

• MCP servers
• Security

Spring AI Agent Utils reimplements Claude Code's core agent tools, including file operations, shell execution, web access, task management and skills, as Spring AI tools for Java AI applications.

• Agent frameworks

Nexus Agent is a Chinese-language enterprise AI agent platform combining conversational chat, document question-answering, web search, retrieval-augmented generation, MCP tool calling and skills extensions behind a layered execution architecture.

• Agent frameworks
• RAG

LLM Tornado is a .NET SDK for building AI agents and workflows against more than 30 model providers and vector databases through one provider-agnostic interface.

• Agent frameworks

yzma is a Go library that embeds llama.cpp directly into Go applications for local inference of language and vision-language models, with hardware acceleration and no CGo dependency.

• Inference

SSH MCP Server gives AI agents controlled, audited SSH access to remote Linux hosts, classifying each command and requiring policy-based or human approval before anything destructive runs.

• MCP servers

hipfire is an AMD GPU-native inference engine, written in Rust, for running large language models locally with an Ollama-style command-line interface and an OpenAI-compatible API.

• Inference

A dashboard that charts GitHub Copilot usage metrics for an organization or enterprise account, pulled from GitHub's Copilot Usage Metrics API, with adoption and impact charts.

• Monitoring

AIGuide is a Chinese-language guide to building AI applications and preparing for AI engineering interviews, covering LLM basics, RAG, agents, MCP, and using tools such as Claude Code and Codex, organized as a learning path with companion agent skills.

• Tutorials

A set of runnable Colab notebooks that teach the applied AI Engineer / Forward Deployed Engineer skill set, including model APIs, RAG, evals, agents and fine-tuning, built from raw API calls instead of a framework, and run on the free Groq API.

• Tutorials

agy-staff is a plugin for Claude Code, Codex and Pi that lets those coding agents delegate research, review and small implementation tasks to Google's Antigravity CLI, giving them a faster secondary model to consult.

• Agent skills

BioMCP is a single command-line binary and MCP server that gives AI agents grounded access to about 30 biomedical data sources such as PubMed, ClinVar, ClinicalTrials.gov, OncoKB and Reactome.

• MCP servers

Awesome Agent Skills is a curated list of Agent Skills, modular instruction packages that give AI coding assistants on-demand capabilities, working across Cursor, Claude Code, GitHub Copilot, Windsurf, Codex, OpenCode and OpenClaw.

• Curated lists

Tura is an open-source agent runtime harness that replaces a typical ReAct agent's repeated model round trips with one runtime-managed command graph, aiming to use fewer tokens while matching or beating baseline success rates on coding benchmarks.

• Agent frameworks