Repository index
6792 AI, ML and Web3 projects · page 14 of 68
- Evolving AI Scientists
Platform for push research with self-training AI scientists automating the scientific process.
- PINVOU-AGENT: AI AGENT FOR DESKTOP
An open desktop AI agent that manages tools, files, knowledge, and workflow, delivering real results.
- Harbor — AI Stack
Stop configuring AI infrastructure — Harbor deploys a full LLM stack with hundreds of services in one command. 3017 stars on GitHub.
- README-AI: AUTOMATIC GENERATOR
AI‑powered tool that instantly creates professional README files for any GitHub repo. Boost productivity and keep docs fresh.
- MCP‑PLUGIN FOR JADX
The plugin adds MCP server support to JADX, enabling AI agents to analyze and modify decompiled code directly.
- 7600+ BRAND SVG LOGOS
Library with 7600+ official SVG brand logos and cloud icons for AWS, Azure, GCP. Access via npm, CDN, CLI and integrations for React, Vue, Svelte, Figma. Free, MIT license.
- AI Presenter from Script
Provider-agnostic Codex skill that creates verified AI-presenter videos from a script and an authorized presenter image. Supports any Codex service and guarantees fidelity to the source material.
- AGENT OS: SPEED AND SAVINGS
Portable open-source OS for agents in Rust and Wasm: cold start ~6 ms, 32× cheaper than sandboxes.
- ULTRACITE - LINTER WITHOUT CONFIG
Opinionated linter and formatter with zero configuration. Just works — right after installation.
- AVO: Generate Go Assembler
The avo project lets you generate x86 assembly code directly from Go programs, simplifying low‑level development.
- AI Token Tracking
🛰️ CLI utility for monitoring token usage in OpenCode, Claude Code, Cursor, Gemini, and other AI tools. 🏅Global leaderboard + 2D/3D chart.
- FAST EMBEDDING SERVING ENGINE IN RUST
Infinity is a high-performance engine for text embeddings, reranking, CLIP, CLAP, and ColPali. Written in Rust for low latency and high throughput. 2963 stars on GitHub.
- VEKSA: AI Meeting Transcription
Open-source API for transcribing meetings in Google Meet, Teams and Zoom with auto‑joining bots and real‑time WebSocket streams.
- GRAPH RAG KNOWLEDGE BASE ON TIDB
autoflow — conversational knowledge base on Graph RAG with TiDB Serverless Vector Storage. Supports CoT and graph connections. Demo: tidb.ai
- Native Switch Games on Apple
Rewrites Switch game code to ARM64 and renders via Metal, letting you play your own Switch library natively on iPhone, iPad, and Mac.
- LANGGRAPHJS TURNS AGENTS INTO GRAPHS
Framework for building reliable AI agents as graphs in TypeScript. Ideal for complex agent pipelines.
- READANY - AI Book Reader
Cross-platform e-book reader with semantic search, RAG chat, and local vector storage.
- BUILDERBOT: CREATE BOT IN MINUTES
Builderbot lets you create a WhatsApp chatbot in TypeScript in minutes. OpenSource project with templates and easy deployment.
- OPEN COMPUTER CONTROL
Open Computer Use – open-source alternative to Codex Computer Use in Swift. It allows controlling the computer via natural language commands, using Claude Code models to access the terminal, files, and macOS apps.
- Claude Extension Hub
A single hub to find Claude Skills, Agents, Commands, Hooks, Plugins, and Marketplace collections to extend Claude Code, Claude Desktop, Agent SDK and OpenClaw
- VORTEX — MOST STRUCTURED DATA FORMAT
Vortex is an extensible framework for columnar data compression and the fastest open columnar format. It is the primary tool for next‑generation analytics from LFAI&Data.
- DIY-LLM: FULL LLM COURSE WITH CODE
The diy-llm project offers six step‑by‑step Jupyter notebooks covering data pre‑training, tokenization, Transformer, MoE, distributed training, scaling laws, inference and LLM alignment. Each assignment progressively increases in difficulty, enabling you to build a full‑stack LLM from scratch.
- J-SPACE COGNITIVE SUITE
J-Space Cognition Suite — tools for managing model output during inference: deep reasoning, long-horizon tasks, verification and recovery. Based on J-space research from Anthropic.
- AI Agent Orchestration Hub
Agent Orchestration Command Center
- AI AGENTS: THE DEFINITIVE GUIDE
A complete guide to building AI agents: from concept to practice in Jupyter Notebook.
- Claude Opus Music Video Generator
JavaScript code that creates a synchronized music video for the track 'I'm Upping My P(doom)' using the Claude Opus 5.5 model.
- MOSS-TTS-Nano: Lightweight TTS
MOSS-TTS-Nano is an open multilingual speech generation model with 0.1 billion parameters. It runs in real time on CPU without a GPU and simplifies deployment for local demos, web services, and lightweight product integration.
- TONS OF SKILLS FOR CLAUDE CODE
tons-of-skills-marketplace — open-source database with 471 plugins, 3069 skills and 347 agents for Claude Code. Available at tonsofskills.com and managed via the ccpi CLI package manager. The project is written in Python and focused on AI‑agent automation.
- UNITY-MCP - AI AGENT FOR UNITY
Unity-MCP adds AI skills and MCP tools to the Unity Engine via a simple CLI, enabling any C# method to become a tool with a single line.
- TRUSTGRAPH: SEMANTIC LAYER FOR ONTOLOGIES
Trustgraph creates a semantic intelligent layer for ontologies, enabling agents to work with a deterministic context graph. It processes documents, extracting structured knowledge and providing explainability for AI decisions.
- OPEN AI FOR SWIFT
The Swift developer community introduced a convenient package for working with the public OpenAI API.
- rl-baselines3-zoo - RL Training with Hyperparameters
Framework for training Stable Baselines3 agents with automatic hyperparameter tuning and ready pre-trained models.
- OSWorld: BENCHMARK FOR AI AGENTS
OSWorld is a benchmark for evaluating multimodal AI agents in real computer environments. Open tasks in real operating systems.
- OPEN ONTOLOGIES AI ENGINE
Rust-based MCP server for creating, validating, and reasoning over RDF/OWL ontologies. Built-in Oxigraph, tabular reasoner, and SHACL.
- VOICEMEM UNIVERSAL MEMORY
VoiceMem provides infrastructure for the next generation of voice agents with universal memory. The left hemisphere stores facts, the right — emotions, and the fully streaming architecture eliminates latency.
- PROJECT MAYA: LOCAL LLM ENGINE
Project Maya lets you run GLM-5.3-Flash (321B MoE) locally on a single GPU or up to 16 GPUs, supporting NVIDIA and AMD on Linux and Windows. The package includes engine, server, dashboard, OpenAI- and Anthropic‑compatible API, and quantization tools.
- FORMAL VERIFICATION OF AI AGENTS
Enterprise agent library with formal verification in Lean. 2850★. Open alternative to costly YC startups. FastAPI + Bayesian inference.
- DeepSeek Reasonix: Terminal AI Agent
Terminal AI coding agent based on DeepSeek. Optimized for stable operation with prefix cache — just run and forget.
- MOSS Image 5B DiT Model
A 5‑billion‑parameter diffusion transformer (DiT) base model for generating high‑quality text‑aware images, supporting layout control and fast fine‑tuning.
- Video Quote Image Creator
Extracts frames with embedded subtitles from videos and generates 3:4 aspect ratio long images for social sharing.
- AI Pentest Assistant
The METATRON project provides a local AI assistant for conducting penetration tests on Linux (Parrot OS), using its own LLM to generate attacks and analyze vulnerabilities.
- EVALSCOPE: SIMPLE LLM TESTING
Flexible Python framework for evaluating the performance of large models (LLM, VLM, AIGC). ⭐2874
- RAMALAMA EASY AI LAUNCH
RamaLama is an open Python tool that lets you quickly run any AI models in containers and use them for inference in production.
- AgentGateway: Proxy for AI Agents
AgentGateway is a new Rust proxy server that simplifies interaction between AI agents and MCP servers. It supports scalable routing and low latency.
- OS FOR AI AGENTS: CONTROL ALL SESSIONS
compozy is an autonomous control plane for AI agents. See and steer every Claude Code, Codex, and Cursor session. Agents check in on their own, you intervene only when needed.
- Synkra AIOS Core
Synkra AIOS: AI-Orchestrated System for Full Stack Development — core framework v4.0, simplifying the creation of full-stack applications with AI integration.
- KAGENT: CLOUD AI AGENT
Kagent — a cloud-native AI agent in Go that simplifies creating and managing AI agents in Kubernetes. It supports Discord integration and fast deployment.
- VOX-DIRECTOR: ONE TOPIC VIDEO
The project turns a single topic into a ready‑to‑use Vox‑style video collage using Atlas Cloud and ffmpeg — a fully automated Python agent‑skill.
- AI Trading Infrastructure
Real-time crypto & stock screening, advanced technical indicators, Bollinger Bands intelligence, candlestick patterns + native Claude Desktop integration. Multi-exchange (Binance, KuCoin, Bybit+), open-source AI trading infrastructure.
- SPIDER: LIGHTNING-FAST WEB CRAWLER
High-speed web data streaming with minimal latency. Ideal for real-time parsing and data collection.
- HYPERFRAMES: AI VIDEO EDITOR
Edit videos, Reels and YouTube Shorts via Codex or Claude Code. 14 skills, transcript‑driven cuts, storytelling, 406 motion‑graphics cards and HyperFrames templates.
- AI Cross-Dressing Image Gallery
1,407 AI-generated cross-dressing reference images with virtual try-on, supporting bilingual search and multiple AI image APIs including OpenAI, Gemini, and Doubao with custom endpoints.
- 🤖 RALPHY - Autonomous AI Agent Development
A Bash script that runs Claude Code, Codex, OpenCode, Cursor, Qwen, and Droid in a loop until your PRD is fully implemented.
- AI for Document Automation
The project uses LLM and vision for OCR, automating document digitization in paperless-ngx.
- CONTROL X64DBG VIA AI
x64dbg‑MCP Server – a native plugin that lets AI assistants control the debugger via HTTP: set breakpoints, step, read memory and registers.
- Agent Skills for Three.js
A set of AI‑agent skills enabling the creation of browser‑based games in Three.js with high‑quality graphics, UI, testing, and optional generation of 3D models, images, and sound.
- REVERSE ENGINEERING TOOLS CATALOG
Collected 464★: tools, cases, and papers for reversing web, games, apps, binaries, firmware, protocols. Covers Android, decompilation, binary analysis. Written in Python.
- Cross-Platform Native UI Framework
Build native apps for iOS, Android, macOS, Windows, Linux, and web using a single Rust codebase with native UI components.
- OM1: Modular AI for Robots
Modular AI runtime for robots in Go. 2808 stars on GitHub!
- ASCILINE: ASCII video engine
ASCILINE is a high-performance engine for real‑time ASCII‑video rendering. It sends binary‑encoded frames over WebSocket and displays them in the browser at 30 FPS.
- SOLON — SPRING ON STEROIDS
Next-generation Java framework: 7x higher concurrency, 10x faster startup, and compatibility with Java 8–25!
- Local AI Hub for WeChat
A system for analyzing WeChat chats with CLI, history search, daily digests, and task tracking. Works entirely locally.
- BRANCHING AND THIN CLONING FOR POSTGRES
DBLab adds branching and thin cloning to any Postgres. Accelerates DB testing in CI/CD, cuts infrastructure costs, and reduces time-to-market.
- AUTODISTILL: TRAINING WITHOUT LABELING
Automatically turns raw images into ready-to-use inference models, using foundation models to generate pseudo‑labels and train supervised models. No manual labeling required.
- GONZO - Lightning Analysis
Powerful TUI log analyzer written in Go: fast viewing, filtering, and visualization of logs.
- ORCALITE LLM ROUTER SELF-HOSTED
Self-hosted LLM router with managed safety net and OpenAI‑compatible API. Supports BYOK, a single workspace, and streaming generation. For advanced routing use the hosted OrcaRouter.
- OpenGap: AI Agents for Git
Framework for defining AI agents via Git — without tying them to a specific platform. A standard built directly into Git.
- Contract-Driven gRPC Test Harness
Snapshots valid gRPC call chains with human-approved baselines so AI agents can verify API regressions without tracing. OpenAPI support planned.
- VibeProxy — AI proxy macOS menu
A macOS menu bar app for using Claude Code and ChatGPT subscriptions with AI tools without API keys.
- MANTIS: TOOL FOR AI AGENTS
A modular skill set for autonomously discovering, reproducing, and fixing vulnerabilities in AI-agent code.
- TEAQL: DETERMINISTIC AI EXECUTION
Makes AI predictable: guarantees the same result on every run.
- CLAUDECODE.NVIM - AI AGENT FOR NEOVIM
Extension for Neovim that integrates Claude Code — a powerful AI agent for writing, refactoring, and debugging code directly in the editor.
- PLUELY HIDDEN AI ASSISTANT
Pluely — open-source Cluely alternative: fast, private AI assistant for meetings and interviews, invisible in video calls and recordings.
- EPHEMERAL WORKSPACES WITH SHARED HISTORY
LayerFS gives each agent a full writable FS over shared history without copying the base. Agents work in parallel, commit changes as deduplicated checkpoints, then branch, rewind or merge. 1964⭐ in Rust.
- MAID: MANAGE AI MODELS
Free tool for working with local and remote LLMs: llama.cpp, Anthropic, DeepSeek, etc.
- Claude Code Framework
Claude Code framework eliminates the need to constantly ask the AI to behave like a senior engineer, saving you time.
- Git-native AI agent protocol
It is a framework‑independent standard that lets you define AI agents directly in a Git repository using plain files and commits. The protocol ensures compatibility with any tools and languages.
- AI analysis of TradingView charts
The project connects Claude Code with TradingView Desktop, enabling automated chart analysis, script creation, and trade execution via natural language.
- CLAUDE BLOG: 30 SKILLS FOR BLOG
A set of 30 subskills and 5 agents for automating a blog on Claude Code. Optimized for Google and AI quotes.
- AX: DSPY FOR TYPESCRIPT
Ax is essentially the 'official' implementation of the DSPy framework for TypeScript, simplifying the creation of LLM applications.
- METAMCP: ONE DOCKER FOR ALL MCP
METAMCP combines MCP aggregator, orchestrator, middleware, and gateway in a single Docker container, simplifying AI agent deployment and management.
- KARPATHY LLM WIKI: BOUNDLESS KNOWLEDGE
Agent Skills‑compatible LLM‑wiki for Claude Code, Cursor, and Codex. Build a Karpathy‑style knowledge base from sources, links, and linting.
- DATACHAIN - DATA WITH AI
Typed versioned datasets for unstructured data on top of S3, GCS, and Azure. A great tool for working with AI information.
- OPEN AGENT SDK WITHOUT CLI
TypeScript SDK for building AI agents without CLI, fully open source and compatible with claude-agent-sdk.
- GPU-Accelerated Manim Animations
An animation engine that renders math and 3D videos using Manim's Python API, accelerated by a Rust GPU backend for blazing speed.
- Lightweight DB Server
Zero-dependency, token‑efficient MCP server for PostgreSQL, MySQL, SQL Server, MariaDB, and SQLite.
- Open-Interface — LLM COMPUTER CONTROL
A Python tool that lets you control any computer using neural networks. Full system control via text commands.
- DOORY — OPEN SOURCE CANVA CLONE
Doory — open-source Canva alternative on TypeScript and Django. Supports real-time collaborative editing via Fabric.js. 1772 stars on GitHub.
- MIND FORGER: NOTES MARKDOWN .
Mindforger — a thinking notebook and Markdown editor in C++. Runs on Linux, macOS and Debian, providing fast search, links and tagging of notes. Supports a knowledge graph and export to PDF, HTML, and plain text.
- KITE Kubernetes AI Dashboard
Modern Kubernetes platform: multi-cluster management, OAuth/RBAC, audit, and AI agents in one workspace.
- SOLANA MCP VYBE: OPEN REGISTRY
Public registry of Solana MCP metadata for Vybe API, simplifying integration and access to blockchain data.
- Universal Content Translator
AI agent for translating audio, video, games, books (PDF, EPUB) and special formats (RPG, ASMR, GALGAME) — one click, high quality.
- HyperQwen: Qwen on Consumer GPUs
The project optimizes running large Qwen models on available GPUs using vLLM patches and requantization. It achieves 127 tokens/s on 24 GB.
- Unified rule manager
Ruler — apply the same rules to all coding agents
- PAPERSGPT — AI for Zotero
Powerful AI plugin for Zotero with support for ChatGPT, Gemini, Claude, DeepSeek and other models. 2459 stars!
- MTPLX SUPERFAST LAUNCH
MTPLX lets you run Qwen 3.8 Flash Next and Qwen 3.8 27B on Mac at 125 tokens/sec in OpenCode. Thanks to MTP speculative decoding on Apple Silicon, results stay accurate at any temperature and server works with OpenAI and Anthropic APIs.
- AGENTS - EXPER SKILLS
Collection of skills for Claude Code and other AI agents, compatible with agentskios.
- LOBE ICONS: AI Logos for React
Ready-to-use SVG/PNG/WebP icons of AI/LLM brands for React and React Native with no dependencies. Just import and use.
- AI Agent Launch Film Maker
Generates commercial-grade product launch videos with music composition and frame-level editing via agent.
- CROSSPASTE DESKTOP: CLIPBOARD SYNC
Crosspaste‑desktop syncs the clipboard between macOS, Windows and Linux over a local network using end‑to‑end encryption without the cloud. It includes OCR text extraction from images, CLI control and an MCP server for extensions.