Repository index
1015 AI, ML and Web3 projects · page 6 of 11
- UNLA — MCP GATEWAY FREE
Unla — lightweight MCP gateway that instantly turns any existing MCP servers and APIs into full MCP servers without changing code. Deployment is via Docker, and management is through an intuitive web UI.
- AI CHATBOT FRAMEWORK ON PYTHON
Framework for creating chatbots with NLU and AI on Python. Supports LLM integration and provides a ready-made dialog manager. Actively developed, already 2166 stars on GitHub.
- OPENBITFUN: AGENT ON RUST
OpenBitFun is a high-performance agent written in Rust, combining the depth of a Code Agent with open, general-purpose capabilities for any tasks. Polished desktop application and scalable agent-teams.
- TOP AI CODING ASSISTANTS
Collected over 2000 stars, a list of the best AI tools for code generation, autocomplete, and refactoring. Supports popular languages and IDE integrations.
- QUICKSC0PE: UNIVERSAL MALWARE ANALYSIS
Universal malware analysis tool. Supports APK, ELF, EXE on Linux. Uses YARA for precise threat detection.
- UNIFIED GATEWAY FOR AI SERVICES
agent-router is an Envoy Gateway‑based gateway that provides unified access to various generative AI services. Implemented in Go, it runs in Kubernetes and adheres to CNCF standards for inference and API gateways.
- GREPAI: AI CODE SEARCH
Grepai indexes the local code via the tree-sitter, providing semantic search and call graph construction. Works offline, 100% suitable for AI agents like Claude Code and Cursor.!!!. Gives access to code without sending data to the network.
- DETIKZIFY DRAWS LATEX DIAGRAMS FAST
DeTikZify turns textual descriptions or hand‑drawn sketches into ready‑to‑use TikZ code for LaTeX. The project leverages HuggingFace and Llama models for inverse graphics generation, saving time on typesetting scientific illustrations.
- APIPARK AI GATEWAY NEW GENERATION
APIPark — gateway for all models: single entry, LLM request management, load balancing and failover. Supports OpenAI, Claude, Gemini, DeepSeek and more, giving unified requests/responses, API applications, call stats and one‑click deploy.
- AI SKILLS FOR SOCIAL MEDIA
AI skills for social media: find outlier posts, mine comments, analyze competitors, ad libraries and trends on TikTok, Instagram, YouTube, Reddit, X, LinkedIn. Works via ScrapeCreators and compatible with Claude Code, Cursor, Codex, Gemini CLI.
- VLLM-METAL: ACCELERATING LLM ON APPLE SILICON
vllm-metal is a community‑maintained plugin that brings Metal‑based acceleration to vLLM on Apple Silicon. It uses MLX to run LLMs on macOS with low latency and high throughput, no CUDA required.
- CLAUDE CODE LEARNS FROM EXPERIENCE
claude-reflect — self-learning system for Claude Code. Captures corrections, positive feedback, and preferences, then syncs them to CLAUDE.md and AGENTS.md. 1690 stars on GitHub.
- COLLECTION OF JAILBREAK LLM METHODS
Awesome-Jailbreak-on-LLMs is a collection of cutting‑edge jailbreak methods for large language models. The repository contains papers, code, datasets, evaluations, and effectiveness analysis.
- GRAPHICAL RAG AGENT FOR COMPLEX QUESTIONS
Controllable‑RAG-Agent is an advanced Retrieval‑Augmented Generation system for answering complex questions. It uses a graph‑based search algorithm and integrates with LangChain/LangGraph, allowing pipeline configuration and source control.
- SPEC‑KITTY: AI DEVELOPMENT
Spec‑Kitty implements Spec‑Driven Development: you write specifications, and AI agents (Claude, Cursor, Gemini, Codex) generate code. Includes a kanban board, git worktrees, and auto‑merge for a smooth workflow.
- EXTRACTTHINKER DOCUMENT INTELLIGENCE
ExtractThinker — an ORM‑style library for LLM integration, including LangChain and LlamaIndex, that streamlines parsing of PDF, DOCX and images, performs layout and analysis and OCR, and provides tools to build flexible document workflows.
- MCP CLIENT TO REMOTE SERVER
mcp-remote connects any MCP‑client that works only via stdio to a remote MCP‑server over the network. It supports bidirectional message exchange, is written in TypeScript and has 1.5k stars.
- VOICE ASSISTANT FOR MAC
Wally lets you control Mac by voice and search local documents offline without internet. It runs on Apple Silicon M1 using an on-device offline LLM via llama.cpp and RAG search. It supports KittenTTS and KokoroTTS for natural speech.
- IMCP MCP SERVER FOR MACOS
iMCP — macOS application that launches the local MCP server to provide access to Messages, Contacts and Reminders via the standard protocol. It enables integration of these services into any MCP‑compatible client.
- DORY: LOCAL ENVIRONMENT FOR APPLE SILICON
Dory provides a full-featured local development environment for Apple Silicon, combining Docker, Compose, Kubernetes and virtual machines. It includes policy‑bound agent sandboxes for secure code execution and AI agents.
- THEPIPE: DATA FROM DOCUMENTS FAST
ThePipe uses vision‑language models to extract data from PDF, Word and PowerPoint. It recognises tables, text and images, turning them into ready‑to‑use structures. Supports batch API.
- REAGENT: AI RECONSTRUCTS CODE
Reagent uses AI to reconstruct and validate C/C++ code from the binaries. Supports analysis of autonomous agents and integration with Claude, Codex.
- OPEN AI-EXECUTIVE TEAM
OpenExecutive — a virtual executive team presented as a single coherent persona, backed by eight specialized Claude agents. The project is built with FastAPI and Next.js, written in Python and already has 1488 stars on GitHub.
- ZERON: MANAGING AI AGENTS
Zeron — a unified control panel for AI coding agents: Claude Code, Codex, Cursor, Devin. It lets you start, stop and monitor agent tasks in real time via CLI and web interface.
- E2B RUNTIME: GO-POWERED CLOUD
Runtime executes code and AI agents in the E2B stack: in the cloud, on-premises, and locally. It is built in Go, uses Firecracker and GCP for isolated sandboxes. It supports Consul for service discovery and working with the code interpreter.
- AntSK Offline Agent .NET 9
AntSK is a knowledge base and agent built on .NET 9 with AntBlazor, Semantic Kernel and Kernel Memory, running fully offline. It supports local Large Language Models and integrates with Aspire for monitoring.
- EDGEVER — AI-ALTERNATIVE TO EVERNOTE
Edgeever — an open‑source AI‑native Evernote alternative with built‑in MCP support. Runs free on Cloudflare (Workers, R2, D1) or self‑hosted via Docker. Lets you create, index, and search notes using LLM.
- XTREME1 — PLATFORM FOR MULTIMODAL ANNOTATION
Xtreme1 — platform for annotating 3D‑LiDAR, images and LLM data. Supports collaborative work, AI suggestions and export to COCO, KITTI, Hugging Face.
- GAWKBOT: AI-TEAM IN TERMINAL
The project lets you quickly spin up a personal office of AI agents working together on a shared context. Supports Claude Code, Codex, OpenClaw, OpenCode and any local LLMs.
- ONE INTERFACE FOR ALL AGENTS
any-agent unifies work with LangGraph, CrewAI, Autogen, etc. Allows switching frameworks without rewriting logic. Supports A2A and MCP for interoperability.
- 48 FREE CLAUDE SKILLS
48 free Claude skills for marketing: Google Ads, Meta Ads, SEO and AI visibility. Simple way to add Ryze MCP via claude mcp add ryze --transport http https://connector.get-ryze.ai/mcp. 1.2k stars on GitHub.
- SMART SECOND BRAIN IN OBSIDIAN
Plugin for Obsidian that connects your private AI‑assistant to your second brain. Lets you ask questions, get answers and save dialogues directly in notes.
- IoT DC3 AI-PLATFORM FOR INDUSTRY
IoT DC3 — an open multi‑protocol cloud-native AI‑platform for industrial IoT. It connects devices, collects data, delivers it from edge to cloud and provides intelligent operations.
- SITE FOR DEVELOPERS AI
🌎💎 site-for-developers is a curated list of AI tools created specifically for developers. The project gathers the best tools for working with LLMs, RAG, multimodal models, and AI agents, updated weekly. It serves as your first bookmark for quick access to the latest AI-stack innovations.
- OPEN DESKTOP-PET WITH AI
OpenPets — a local desktop companion platform with animated pets, plugin SDK, and coding‑agent integrations (such as Claude Code). Lets you create your own pets and extend their behavior via TypeScript plugins.
- TEMPORARY WORKSPACES FOR HUMANS AND AI AGENTS
Free4Chat — platform for temporary interaction between humans and AI agents. Lightweight architecture on Cloudflare Workers. Create rooms, share files, and connect agents in seconds.
- AI PROJECTS ON AUTOGEN AND TUTORIALS
The repository contains a set of examples and tutorials on using AutoGen to create AI agents. Here you will find ready‑to‑run Python scripts, integrations with CrewAI, LangChain, LM Studio, and SmolAgents.
- AWS SKILLS FOR AI AGENTS
Ready-made skills for integrating AI agents with AWS services. Supports Claude Code and Codex. 1155 stars on GitHub.
- OKOU — AI AGENT FOR TEAMS
Okou connects to your team's tools and executes tasks across marketing, sales, engineering, and operations. Runs agents in isolated sandbox with access control. Written in TypeScript, 1.1k stars on GitHub.
- OKOU AGENT FOR TEAMS
Okou integrates with your team's tools and autonomically performs tasks in marketing, sales, engineering, and operations. You retain full control over its actions via a simple interface.
- ELASTIC LABS: EXAMPLES FOR SEARCH
The repository contains notebooks and example applications for search, observability, and security based on Elasticsearch. It includes ready‑to‑use solutions for AI applications, chatbots, and log analysis.
- BEAUTIFUL CLAUDE CODE CHAT FOR VS CODE
VS Code extension adds elegant chat interface for Claude Code. Supports dialog history, syntax highlighting, and convenient code navigation. 1087 stars on GitHub.
- UNIFIED ADS AND GA4 MCP SERVER
The server combines Google Ads MCP, Meta Ads MCP, and GA4 into a single endpoint, offering 250+ tools for Claude, ChatGPT, Cursor, and n8n. It runs as a remote MCP with OAuth login, requiring no API keys. Built by Ryze AI.
- SUBTITLES FOR ENTIRE SEASONS IN MINUTES
Translates whole season subtitles in one pass. Supports .srt/.ass/.vtt/.lrc, 120+ languages, 27 LLM providers. Timing untouched!
- MARKLLM: WATERMARK FOR LLM
MarkLLM — open-source toolkit for watermarks of large language models. Allows embedding, checking and comparing different watermark schemes. Supports popular LLM and integrates via a simple API.
- FAST ALTERNATIVE FIRECRAWL IN RUST
crw — fast Firecrawl/Tavily alternative in Rust: web scraper, crawler and search API with MCP server for AI agents. Compatible API (/scrape, /crawl, /search), runs 2.3× faster than Tavily and 1.5× faster than Firecrawl on 1 000 URL, uses 6 MB RAM, binary, hosting.
- SEARCHAT AI SEARCH CHAT
SearChat combines AI chat and search in a single interface. Supports OpenAI, Anthropic, VertexAI, Gemini, DeepResearch and the SearXNG aggregator. Available as a Docker image for quick deployment.
- CHROMEM-GO: VECTOR DB WITH ZERO DEPENDENCIES
Embeddable vector DB in pure Go. Chroma-like API, zero external deps, runs in-memory with optional persistence.
- DR-CLAW: AI‑LABORATORY OF SCIENTIFIC RESEARCH
dr-claw is an AI‑laboratory where dozens of AI assistants work like research doctors. The platform unites Claude, GPT‑4 and other models in a single IDE for automatic planning, execution and documentation of scientific tasks.
- GPT-6 ASTRA: PROMPTS INTEGRATIONS
Evidence-backed use cases, prompts, integrations, evaluations, and safety notes for OpenAI GPT-6 Astra.
- GLM-5.3 FLASH AND J-SPACE BENCHMARK
The project shows integration of GLM-5.3 Flash with J-Space for full benchmarking the J-Space Cognition Suite. Results show agent skills implementation and baseline comparison.
- JONEX UNIVERSAL MULTIMODAL PARSER
Jonex — universal parsing engine for multimedia, text and code, extended with ontology and LLM‑wiki. It indexes data, builds a knowledge graph and provides ready‑to‑use prompts for LLM agents.
- OPEN TEMPLATE AI AGENT
Template for building a file-system and knowledge-based agent in TypeScript. Lets you create an AI agent that stays up-to-date with your knowledge base. Includes integrations with Nuxt and RAG support.
- FAST FASTEMBED-RS IN RUST
fastembed-rs — a Rust library for local creation of vector embeddings and their reranking. It has 1009 stars on GitHub, it is suitable for RAG, search and rerank models.
- AGENT API CATALOG
awesome-agent-apis collects over 660 generative media models hosted on muapi, plus community‑submitted API tools for SEO, enrichment, social media, and web scraping. Each item is described by a single YAML file, easy to filter by capability. The project has 1008 stars and is written in Python.
- ONTOGPT: ONTOLOGY EXTRACTION
Ontogpt — an LLM‑tool for extracting ontologies from text, including the SPIRES method. It builds hierarchical taxonomies and generates OWL/RDF schemas ready for use in data modeling.
- OPEN BEDROCK GATEWAY API
The project gives an OpenAI‑compatible REST interface to Amazon Bedrock, allowing models to be called via familiar requests. It supports chat/completions, embeddings endpoints and text generation. It runs on Python as a lightweight proxy.
- TURN YOUTUBE INTO SHORTS IN MINUTES
Windows app turns long YouTube videos into vertical 9:16 clips. AI detects best moments, tracks face, and adds word-by-word captions.
- BANANA-CLAUDE: AI IMAGE GENERATION IN CLAUDE
Banana-claude adds image generation to Claude Code via Gemini. Write a prompt — get a ready asset without leaving the terminal. Open-source skill for creative tasks.
- MCP SERVER FOR NEO4J IN PYTHON
MCP server from Neo4j Labs implements the Model Context Protocol for Neo4j via stdio interface. Allows LLM agents to run Cypher queries and get responses. Supports asynchronous operations and easily integrates into Python projects.
- MEMORY CLAUDE CODE IN OBSIDIAN
The project reduces tokens in Claude Code session by 71,5× via Obsidian + Graphify. It provides persistent memory, knowledge graphs of the codebase and chat import. Interface in Portuguese (PT‑BR).
- GROUNDING LMM: SEE AND UNDERSTAND
CVPR 2024 🔥 Grounding Large Multimodal Model (GLaMM) – the first model that generates natural language answers tied to object segmentation masks. It combines scene understanding with precise textual explanations.
- UNIVERSAL CONTENT READER
x-reader — MCP server that reads content from more than 10 platforms: social networks, blogs, repositories and documents. Written in Python, it supports text and metadata extraction and provides a ready‑to‑use pipeline for AI agents.
- MCP PERMANENT NOTES
Local note-taking system for Model Context Protocol with persistent sessions, recording decisions and evidence. Supports resumable tasks and storage in SQLite.
- COMPREHENSIVE CODE SEARCH LOCALLY
zvec-grep — local search across your workspace, using BM25 and tree-sitter for instant indexing. Supports CLI and integration with AI‑agents, letting you search code, documents and configs without sending data over the network.
- EYES FOR AI AGENTS IN PDF
citra gives AI agents vision for PDFs: extracts text, tables, OCR, and visual evidence. Runs locally on Rust, returns page-level citations via MCP. Perfect for RAG and document intelligence.
- DEEPGIT: GITHUB HUNTER
DeepGit — a Python agent for finding the best GitHub repositories by topic. It uses LangChain and LangGraph to perform deep code and documentation analysis. Returns a sorted list with relevance scores and direct links.
- OPEN OPENAI WORKSHOP
Workshop materials for building intelligent solutions with OpenAI on Azure. They provide notebooks, Cognitive Services integration examples, RAG pipelines, and fine-tuning scripts. Perfect for starting Python project quickly.
- ANYGPT — UNIFIED MULTIMODAL LANGUAGE MODEL
AnyGPT is a universal multimodal model built on discrete sequence modeling. It generates images, music, and speech from text and understands translation between modalities.
- LLM-LSP SERVER TO CODE COMPLETION
llm-ls — an LSP server in Rust that connects LLMs to any editor for code completion and generation. It supports local models via llama.cpp and remote APIs such as HuggingFace and OpenAI, using tree-sitter for precise and useful aid context.!!!
- RAPYD ADMIN — LIGHT LARAVEL PANEL
Rapyd Admin — an open-source admin panel for Laravel, created as a lightweight alternative to Nova, Filament and Backpack. It uses Livewire and Blade, allowing you to quickly build CRUD interfaces without extra code.
- ZERO-TO-SGLANG: LLM INFERENCE TRAINING
The Datawhale and SGLang course teaches LLM inference from scratch: build a mini‑SGLang, study the source code, and land your first PR. Available in English and Chinese.
- OMEGA‑AI JAVA FRAMEWORK FOR DEEP LEARNING
Omega-AI - Java framework for deep learning with automatic differentiation, multithreading, and GPU acceleration via CUDA and cuDNN. Enables fast building and training of YOLO, diffusion and LLM networks for inference and training.
- LLaVA++: NEW LEVEL OF MULTIMODALITY
LLaVA++ adds support for Phi‑3 and LLaMA‑3 models to the base LLaVA, enabling multimodal chat and image generation. Users gain better text‑and‑image understanding thanks to integration of cutting‑edge LLMs.
- OPEN AI TEAM
OpenSwarm — orchestrator of autonomous AI‑team based on Claude Code CLI. Control via Discord, sync tasks with Linear and use cognitive memory for context.
- FREE AI CHAT LOG EXTRACTOR TOOL
Extracts chat histories from Claude Code, Cursor, Windsurf, Aider and Cline/Roo Code, saving data as JSON and CSV for analysis and sources. Works via CLI with batch processing support.
- GOLF: ADVANCED FRAMEWORK FOR AI
Production-ready MCP server framework for building, deploying and scaling the AI-agent infrastructure. Includes auth, observability, debugger, telemetry and runtime. Enables running real-world MCPs that power AI agents.
- AGENT SKILLS FOR MOBILE SCREENS
The appllama-skills project provides a set of agent skills that turn top‑app templates into native mobile screens. It works with Claude Code and enables generating UI‑code directly from pattern descriptions.
- VTCODE: SMART AI AGENT
VTCode — an open-source Rust agent for the terminal that writes and edits code directly from the CLI. It supports plugins and skills for integration with LLMs.
- GENTLE SHELL AI ARCHITECT
Gentle‑shell turns Raspberry Pi into a powerful harness for senior architects: SDD/OpenSpec, subagents, strict TDD evidence, and review guardrails. It automatically discovers agent skills and ensures repeatable code quality.
- SMART NOTIFICATIONS FOR CLIENTS
Cross‑platform tool agent‑notifications delivers smart notifications for Claude Code and Codex CLI: pop‑up alerts, sound signals, click‑focus and webhooks. Works on macOS, Linux and Windows.
- ADORABLE — AI AGENT FOR CODE
Adorable — open source Lovable, generating code via AI agents. Supports TypeScript and integration with popular LLMs. Enables rapid creation of agents for code generation.
- CLODDSBOT – AI TRADING AGENT
Open Source AI trading agent operating autonomously on 1000+ markets, (Polymarket, Kalshi, Binance, Hyperliquid, Solana DEX, 5 EVM chains). Scans the market for the edge, instantly executes trades, manages risk and supports M2M payments.
- ONE-API-PRO: ENTERPRISE AI GATEWAY
one-api-pro — enterprise‑grade AI API Gateway based on one-api, with a complete architecture redesign. Added subscription plans, active-active clusters and precise billing. Outperforms oneapi and newapi in functionality and reliability.
- GPT-VIS: AI DATA VISUALIZATION
🤖 GPT-Vis — data visualization for AI Friendly! AntV's entry for AI, 100% generated by AI. Explore data in a convenient visual format.
- MOBILEGYM: ANDROID AGENT TRAINING
MobileGym is a browser‑hosted Android emulator for research on mobile GUI agents. The platform provides verifiable evaluation and highly parallel online reinforcement‑learning training.
- CHETAHCLAWS: AGENT HARNESS FOR AI
CheetahClaws — a fast and easy-to-use infrastructure for agents, supporting long-horizon tasks, multiple models, and tool usage. Written in Python, it has already gathered 775 stars on GitHub.
- SPECTRA-APP: SPECIFICATION FOR AI AGENTS
Spectra-app is a desktop app and CLI that turn specifications into working code for AI agents. Supports Codex, Claude Code, Cursor, Copilot, Antigravity and Junie, accelerating development via a specification‑driven approach.
- INTERACTIVE NPCS WITH LLM
The Interactive LLM Powered NPCs project provides open-source Rust code, allowing you to embed LLM-controlled NPCs in any game for dynamic dialogue and behavior.
- CLAUDE CODE JAVA: READY INFRASTRUCTURE
claude-code-java — reusable AI development infrastructure for Java, tailored for Claude Code. It includes projects templates, CLI build scripts, and abilities for refactoring, commenting, and unit testing.
- IDASSIST: AI ASSISTANT FOR IDA PRO
IDAssist — a plugin for IDA Pro that uses large language models to automate reverse engineering. It generates comments, renames functions, and restores data structures directly in the IDA database.
- REVERSE CODE GENERATION EF
The generator creates POCO classes Code First from an existing DB, exactly like manual reverse engineering. Code fully customizable via T4 templates. Free for academics; commercial license required.
- SEARCHGPT: AI SEARCH WITH SOURCES
SearchGPT — an LLM‑based search engine that returns answers with source citations. It supports web and local file search, providing verifiable information. Works via OpenAI API and integrates easily into projects.
- JARVIS: VOICE-FIRST AI ASSISTANT FOR macOS
JARVIS — a voice‑first AI assistant for macOS, inspired by Marvel. Speak to your computer, it responds and performs tasks: launching apps, controlling the system, searching for information.
- MOSS: INSTANT SEARCH FOR AI
Moss — is a retrieval layer for production AI systems, providing search under 10ms without vector databases. Works in browser, on edge, on-device, and in cloud, written in Python.
- RANKIFY: TOOL FOR RAG
Rankify — Python toolkit for search, reranking, and RAG. Includes 40 benchmark datasets, 7+ search methods and 24+ ranking models. Supports multiple RAG schemes and easy integration into AI projects.
- XYNE: AI SEARCH FOR WORK
Xyne — open-source AI search & answer engine for work, alternative to Glean. Written in TypeScript, it uses Bun, Claude and AWS Bedrock for fast indexing and answer generation.
- DISTILL: LIGHTWEIGHT CODING AGENT WITH TUI
Distill is a lightweight coding agent with a terminal interface, written in Rust. It enables achieving greater results by using significantly fewer tokens when working with LLM models.
- PEBREL AI TERMINAL ON GPU FOR WINDOWS
pebrel — AI-native terminal for Windows, GPU-accelerated via DirectX. Supports SSH, persistent sessions, split panes and deep integration with AI-CLI (Claude Code, Codex).
- RELEVANT PAPERS ON REASONING LLM
Collection of the latest papers and benchmarks on reasoning with foundation models. Enables quick discovery of cutting-edge research and comparison of their performance. Supports multimodal and agentic approaches.