Repository index
6791 AI, ML and Web3 projects · page 65 of 68
- Professional AI Chatbot
Multifunctional chat with 7+ AI-API support (DeepSeek, GPT, Claude, Gemini, Gemma, Qwen) and integration with VS Code / Cursor AI.
- Kohya-LoRA-Tool for Windows
Simple desktop tool for training LoRA models on Windows: supports styles, characters, and video, compatible with SD1.5, SDXL, FLUX and other models, includes built-in download and auto-update.
- Scientific skills for AI agents
The skill library turns any AI agent into a scientist: 165 ready-made skills and access to 100+ scientific databases.
- 1Panel AI Gateway
1Panel AI Gateway is a comprehensive solution for managing AI services: a single entry point, intelligent routing, and full compliance auditing, enabling secure, efficient, and controlled AI deployment in enterprise.
- FL Studio 25 Producer
FL Studio 25 Producer Suite (2026) provides an optimized audio engine, offline licensing, and features to unlock all plugins, including AI stem separation and ultra-low ASIO latency.
- Video Diffusion Restoration
The project offers a spatiotemporal diffusion model for video: HD restoration, watermark removal, and semantic editing in a unified framework with task-adaptive adapters and guaranteed frame stability.
- AI Agent Gateway
Provides a Go‑based gateway that secures, governs, and observes MCP tool calls and LLM traffic for AI agents.
- Live Page Verification Tool
Verify deployed pages against requirements with SHIP/FIX/CAN'T SEE decisions, then patch only identified gaps.
- Multi-Provider LLM Chat Module
Integrates local and cloud LLMs into JainaCore 3.3.5a with semantic caching for efficient AI-driven NPC dialogue and chat.
- IDM 6.43 Patched Downloader
Patched Internet Download Manager 6.43 Build 11 with SOCKS5 support, 403 resume, and MP4 grabbing.
- Heroes Olden Era Data Explorer
Interactive viewer for Heroes of Might and Magic: Olden Era game data including units, heroes, spells, and artifacts with 3D model exploration.
- Type-safe API statuses
A unified pattern for handling API responses in TypeScript with full type inference, reusable instances, and client adapters.
- Image to Code Conversion
The project uses the Codex model to reconstruct images up to 750 px wide into clean HTML/CSS code and performs precise slicing of transparent PNG files.
- Reset Claude Device IDs
PowerShell script that clears local Claude device identifiers on Windows while preserving chat history. Run to reset device tokens without losing conversations.
- Modular Multi-Agent OS with Telegram
Python-based autonomous multi-agent operating system featuring a Telegram command center for human-in-the-loop control.
- AI Meeting Notetaker for Teams & Meet
Joins Teams and Google Meet calls, transcribes Hindi, Telugu, and English mixed speech, then emails English summaries with decisions, action items, and attendance. Telegram check-in controls bot participation.
- FiveM Enhanced Roleplay Framework
A complete 36-module server-authoritative RP framework for FiveM Enhanced with 22 in-game editable content domains and one-line plugin integration.
- Retro LOVE Game Collection
A collection of simple retro-style games built with the LOVE 2D framework, playable in browser.
- Telegram Productivity Bot
Multi-feature Telegram bot with Pomodoro timer, image-to-PDF and PDF-to-image conversion, hosted on Render.
- Telegram Auto-Forward Bot
Automatically forwards messages between Telegram chats using configurable rules for channels, groups, and users without writing code.
- Immortal Tamed Pets
Makes tamed animals immortal in Minecraft by applying creative-mode invulnerability. Works with vanilla and modded tamable entities.
- Warface Data Path Manager
A cross-platform desktop tool to locate Warface game data paths, manage configuration files, and export caches and folders.
- Autonomous Experiment Orchestrator
Orchestrates versioned research-plan pipelines with DAG execution, checkpointed resumable runs, UCB1/Thompson scheduling, provenance ledgers, and offline report synthesis.
- Hacker News Video Player for TV
Plays Hacker News-recommended videos through the YouTube app on Android TV devices.
- Codex Plugin Templates
The project provides ready‑made templates for creating Codex plugins tailored to specific developer roles, simplifying code generation and workflow integration.
- Mosaic Restoration with MiniMax
Restores mosaicked images using a MiniMax H3 diffusion model fine‑tuned with custom LoRA adapters.
- 🐍 AUTORESEARCH: AI AGENTS TRAIN LLM
A system where AI agents autonomously run experiments to train language models on a single GPU.
- Academic Writing Enhancer
Provides model‑agnostic editing of academic texts in Chinese and English, preserving author voice and evidence while reducing formulaic prose.
- AlumNet — Alumni Platform
AI platform for an intelligent exhibition of school history and alumni management with a 'Star Map of Graduates'.
- Ghidra RPC Agent
The project provides an RPC interface for interacting with Ghidra, allowing the creation of Python reverse engineering agent skills.
- LoRa Training for YuE
Train LoRa adapters for YuE models directly inside ComfyUI using Python.
- Claude Code Harness Visualizer
Maps Claude Code harness components—files, memory, tools, permissions, hooks—as diagrams organized by folder and spawn method.
- Local AI Video Studio
Open-source video creation suite running locally on ComfyUI. Train characters, direct 3D scenes, sketch-to-video, and restore footage using only open-weight models.
- Trend2Video-Pro: Video Agent
Trend2Video-Pro is an open-source framework that turns trends into finished videos for creators. AI agents find topics, write scripts, and assemble a clip.
- AGENT SPHERE: AI ORCHESTRATION
Platform for orchestrating AI agents with an LLM engine, built-in tools, and support for MCP, CLI, and browser operations.
- ArachneC2: P2P C2 framework
ArachneC2 is a decentralized libp2p-based bot control system, ensuring secure communication and flexible extensibility via modules. It supports AI traffic analysis for anomaly detection.
- AuditMind: AI Audit Agent
AuditMind is an AI-agent system for managing regulatory knowledge, extracting traceable rules, auditing documents, and intelligent question-answering search over sources.
- AI Meeting Knowledge Workspace
OpsPilot‑AI is an open‑source TypeScript workspace that uses AI to capture meetings, organize knowledge, track tasks and risks, and maintain organizational memory.
- AI Prompt Optimization Framework
A universal prompt template that reduces token usage, enforces security, and maximizes output quality across AI models.
- LLM Prefix Cache Simulator
Simulates prefix caching mechanisms for LLM inference to analyze cache hit rates and memory efficiency.
- AI Agent for CLI Code
CLI utility that uses Gemini 3.5 Flash for automatic code generation, workspace setup, and terminal assistance on Windows, macOS, and Linux.
- RLBot AI Bot Framework
A complete set of tools and templates for creating, testing, and deploying high-performance AI bots in the RLBot ecosystem in C#.
- AI Image and Video Generator
The platform provides access to uncensored AI models to create photorealistic images and videos without unnecessary restrictions.
- OF Content Downloader
A safe and ethical tool for local content archiving and personal data backup.
- Modern TF2 Mod Framework
The codebase simplifies creating Team Fortress 2 mods by offering ready-made tools and minimal boilerplate code. It suits both beginners and experienced developers.
- Stable Diffusion WebUI
Optimized web interface with Forge support, extensions, and guides — everything for working with Stable Diffusion.
- Kairos: Quant Pipeline
Data-driven system for researching and simulating quantitative trading. Supports statistical hypothesis testing, optimization through Polars/DuckDB, and market state classification with ML.
- WHOOP Offline Companion
The app connects a WHOOP band via Bluetooth and saves all data locally on the device without cloud, account, or subscription.
- Crypto framework for collaborative applications
Rust library providing primitives for creating collaborative applications that work over an untrusted server using modern cryptography.
- PH-Content Content Management
PH-Content lets you manage and deliver customizable content, improving user engagement and optimizing workflows.
- Local AI Generator
Fully autonomous offline image-generation studio based on Stable Diffusion for Windows. Automatically configures CUDA for Nvidia and Vulkan for AMD/Intel Arc, requiring no system dependencies.
- Science in Real Time
The project provides interactive real-time science tutoring where H3 Max video is generated continuously and can be interrupted by the student to ask questions.
- AI Evaluation Framework
A library for comparing models by accuracy, p95 latency, and cost, providing an evaluation of each variable based on reference data.
- Streaming Farsi ASR Fine-Tuning
Fine-tunes NVIDIA Nemotron 3.5 ASR for streaming Persian speech with data prep, training, evaluation, and inference scripts.
- Analytic Context Layer
ClariLayer preserves your analytic context between AI sessions, automatically loading the needed data and metadata via the MCP protocol.
- 125B MoE Inference on RTX
Run Qwen3.8-Flash-Next (125B MoE) with 1M context on one or two RTX 3080 10GB cards. Tuned, benchmarked fork of Strata.
- Passive Meshtastic LoRa Reception
Wideband passive interception of Meshtastic LoRa with multistation fusion and offline PSK recovery.
- Training AI to Launch Meta Ads
The project shows how to train an AI agent to automate the full Meta ads workflow — from API setup to campaign launch, browser automation, and precise conversion tracking.
- SparklingKit: Local AI Platform
SparklingKit — a local toolkit for OCR, transcription, translation, image generation, and mind‑map creation, enabling composable workflows on DGX Spark.
- OAuth Flow Visualizer
Interactive HTML map visualizing OAuth, Conditional Access, Entra Agent ID, and MCP authentication flows step‑by‑step.
- DGX Spark Monitoring Dashboard
Provides a read-only web dashboard and rack panel to monitor DGX Spark, ASUS GX10, and MSI EdgeXpert nodes running vLLM or SGLang.
- 🤖 Skills for Engineers
A collection of skills for Claude from the .claude directory by Matt Pocock. Automates routine tasks and tailors AI to your workflow.
- QuantumFlow: Distributed LLM Inference
A distributed framework for LLM inference scheduling with support for multiple backends and adaptive strategies.
- Obsidian Control Center
Windows application for managing Obsidian plugins, themes, tasks, and workspace resources from a single interface.
- JEV DataOps Workbench
Streams data, evaluates quality, auto‑trains LoRA, and evaluates held‑out models.
- AI for Designers
Converts screenshots into React, Figma, and Tailwind CSS components using Claude Code.
- AWESOME-AGENT-EVOLUTION: AGENT EVOLUTION MAP
Open map covering the evolution of AI agents, their self-improvement, memory, skills, and swarm systems. 🌟 208 stars, JS
- Ontological E-commerce Analytics
The system uses ontologies for integration and semantic analysis of e‑commerce data, enabling the discovery of hidden sales patterns and the recommendation of optimal strategies.
- Mixture-of-Experts LLM Engine
Runs an 80‑billion‑parameter mixture‑of‑experts language model on Apple Silicon, fitting within 16 GB RAM.
- Rust Inference Engine for AMD R9700
A Rust-based inference engine optimized for AMD Radeon RX 9700 series GPUs, enabling hardware-accelerated ML workloads.
- Vanguard Vision for Valorant
Vanguard Vision (2026) is an external computer vision framework for Valorant featuring ultra-low latency color detection, mouse automation via Arduino/HID, and a secure kernel-mode overlay.
- Dataroom — AI Catalog
The project creates a personal dataroom from a natural‑language query, using a self‑contained Qwen3.6 on a single NVIDIA L4 GPU for accelerated data exploration.
- Financial data skills
The project provides a set of Python modules for obtaining, processing, and using market data in Claude Code and other AI agents.
- AI Remote Computer Access
Enables ChatGPT, Claude, and MCP clients to securely control a remote computer via natural language commands.
- Phone Line Over USB-C
Open-source 4-layer KiCad board for LTE data, SMS, and voice calls over USB-C with audio routing.
- Agent-Driven Digital Human Video Generator
Python framework for agent-driven continuous digital human video generation with material reuse, multi-style editing, and lip-sync locking.
- FL Studio 25 Producer
FL Studio 25 Producer Suite (2026) offers advanced audio‑engine optimization, standalone licensing, and access to the full plugin set, including AI stem separation and ultra‑low latency ASIO.
- NovoWeave: Generative Design
NovoWeave is a conceptual pseudocode implementation of a framework for generative protein design using deep learning methods. It demonstrates a VAE architecture and approaches for evaluating generated sequences.
- Fast In-Memory Vector Search Engine
A Java 21 Loom-based engine delivering sub‑millisecond KNN search with HNSW graphs and 8‑bit quantization.
- Local AI Assistant for Windows
Run a private AI with chat, vision, RAG, web search, and voice entirely on your Windows PC — one-line install, auto-tuned models.
- AgentPet for macOS
AgentPet is a native macOS app that displays a pet in the menu bar and on the desktop while monitoring the real‑time activity of AI agents Claude Code, Codex, and Gemini CLI.
- KISS-IMU: Odometry Learning
Self-supervised inertial odometry with motion balancing and uncertainty estimation. Finalist at ICRA'26.
- RAG to Skills for AI
The tool converts Retrieval-Augmented Generation into autonomous skills for AI agents, simplifying RAG integration into workflows.
- SEMBLE: Code Search for AI Agents
Hybrid BM25 + semantic search with AST and dependency analysis. A replacement for grep/cat/read/ls for Claude Code, Cursor, Aider, and other AI agents.
- Full AI Agents Course
A 28-chapter practical course on building AI agents: from ReAct to Claude Code, MCP, RAG, and DSPy. All examples are executable Python code.
- Simple TypeScript Video Editor
A lightweight video editing tool built in TypeScript that lets users trim, merge, and apply basic effects to video clips directly in the browser.
- Claude Skills for Indian Trading
Python-based Claude Code skills for Indian market analysis including NSE/BSE data, F&O strategies, technical charts, fundamental screening, and tax compliance.
- Local LLM proxy
Local proxy for connecting OpenAI Codex CLI to popular LLMs: MiMo, DeepSeek, Qwen, and others.
- AI Assistant ZJunCher
AI assistant based on RAG with hybrid search and multi-round memory for team knowledge work and learning.
- World-Scale ATProto PDS
VLPDS is a Rust implementation of an ATProto personal data server designed to handle world‑scale workloads with high efficiency and resilience.
- MiniMax-H3 HyperFlow Integration
ComfyUI node pack implementing HyperFlow's 8-step LoRA for MiniMax-H3 video and audio generation without core modifications.
- Zero-Trust MCP for IDE
A unified Zero-Trust MCP server with semantic codebase search, isolated project memory, and hallucination‑free RAG.
- Crypto Strategy Backtester
Contains Python scripts implementing various crypto trading strategies, each validated through a seven‑stage process on six months of historical data.
- MOVIE CATALOG OF 11 AIRLINES
The project collects monthly movies from 11 airlines' entertainment systems, including British Airways and Emirates. The database includes IMDb and TMDB identifiers. An agent skill picks movies for your flight.
- On-Device Voice Dictation
Local voice dictation using Whisper/Parakeet and a LoRA-tuned Qwen model to correct transcripts on-device.
- Rose — PyTorch Optimizer
Stateless PyTorch optimizer that normalizes gradients for stable and balanced model training.
- Code Problem Search
Lifuyue/issue-finder — a local tool for preparing task handoffs to developers using AI agents. It analyzes code, identifies issues, and suggests specific steps to fix them.
- UniRL: Multimodal Learning
UniRL is a Python framework for unified reinforcement learning of multimedia models. It enables joint training of vision, language, and audio in a single RL cycle.
- C# + Rust UI Framework
An experimental framework for creating user interfaces in C# that uses bindings to the Rust GPUI library for GPU‑accelerated rendering.
- X Ambient Hover Light
Shows ambient light effects from hovered timeline photos and videos across the whole page, reflecting colors from media. Works on X timeline.