Repository index
6791 AI, ML and Web3 projects · page 62 of 68
- SHP 10X FASTER FEATURES
SHP accelerates feature delivery 10x, allowing tokens to be burned at scale the right way.
- LITER-LLM UNIVERSAL LLM CLIENT
The client supports 142+ providers and 11 native libraries, built on a fast Rust core.
- MYTHOS-ROUTER: LOCAL AI-AGENT
Local router with a writing discipline for precise coding via LLM.
- MIRA: AI Workspace for Investors
Agent-native workspace for investment research with evidence tracking and updatable theses on stocks, earnings, and macro.
- WAVEFRONT AI MIDDLEWARE REVOLUTION
Wavefront — enterprise AI middleware, an alternative to UnifyApps, n8n, and Lyzr in Python. Flexible orchestration and quick start for AI solutions.
- AI-DIAL SDK: Build AI Apps
Framework for quickly building apps and model adapters for AI DIAL in Python. Supports integration with LLM and RAG.
- Core Framework CSS Generator
Visual CSS framework and design-token builder for the web, WordPress, and Figma, allowing style creation without writing code.
- MCP ON .NET: SERVER AND CLIENT SAMPLES
Collection of 201★ samples for Model Context Protocol in C#. Covers building MCP servers and clients, LLM integration, and context passing. Ready templates for quick start.
- PI SESSION MANAGER FOR AI CODING
Pi Session Manager lets you browse, search, and resume Pi AI coding sessions. Built with TypeScript and Rust, React for UI. 201 stars on GitHub.
- DEFENSE MEMORY AI-AGENTS
Provides runtime defense against memory AI-agent poisoning (OWASP ASI06). Drop-in for LangChain, OpenAI Agents, AutoGen, CrewAI and mem0.
- LOCAL MCP FOR LOGIC PRO
Local MCP server for Logic Pro control with state persistence and fail-closed protection. Reads project in real time, written in Swift.
- AI Scanner for Warzone
AI framework for pixel scanning and mod menus in COD Black Ops 6. 202 stars.
- HIVE: AI AGENTS IN BROWSER
Browser-native hive-mind combines Claude Code, Codex, Gemini, and OpenCode in one PTY window for collaborative coding.
- AI AGENTS HANDBOOK ON GITHUB
Free guide to building AI agents: from architectures to prompt injection protection. 202 stars 🔥
- HERMES AGENT CN, 202 STARS!
Desktop app for Hermes Agent CN on Windows to use an isolated agent core. Built on Tauri, TypeScript, and Rust.
- FLOWCRAFT - AI AGENTS IN GO
Production-grade Go SDK for building AI agents with long-term memory, knowledge search, and voice control.
- SWIVAL CODING AGENT CLI
Small but powerful CLI agent for writing code, working with open LLM models. Speeds up development without dependence on commercial services.
- KELOS: AI Agent Orchestration
Kelos allows managing autonomous AI agents in Kubernetes, accelerating code writing.
- AGENVOY — Your AI Assistant
Full-featured AI assistant in Go for task automation. Already 202 stars! #likeHub #development
- MGM-OMNI: FUTURE AI VOICE
MGM-Omni enables scaling large language models for personalized long-form speech, ensuring coherent synthesis for hours ahead.
- GENERATE AI WORKFLOW FROM TEXT
FlowX is an MCP server and Python SDK to generate, run, and refine agent workflows from natural language. Supports A2A and FastAPI.
- DEEPSEEK-CODE-WHALE - AI CODING AGENT
Terminal AI agent for DeepSeek with MCP integration, skill system, and cache optimization for a 97% hit rate.
- CODEX-APP-TRANSFER: API BRIDGE
Local desktop gateway that translates OpenAI Codex CLI's Responses API to Chat Completions for Kimi, DeepSeek, GLM and other compatible providers.
- ONIONCLAW: TOR FOR AI AGENTS
Provides AI agents full access to the Tor network and dark web data without configuration.
- NODE9-PROXY: SHIELD FOR AI AGENTS
node9-proxy provides deterministic management and auditing of autonomous AI agents, ensuring safe code execution.
- REPOPROMPT-CE — SIMPLE AI CODE AGENT
Reprompt-ce — community edition of a native macOS app for context engineering of AI agents, including MCP CLI.
- RSSH: Reliable SSH Client
Cross-platform SSH client in Rust that leaves no traces on the system and works fast and reliably.
- ONGRID: INFRASTRUCTURE AGENT
Ongrid – an AI agent that analyzes infrastructure, finds the root cause, and fixes it directly from Slack, Telegram, Lark, or DingTalk.
- META_KIM: SMART ASSISTANT
Governed execution layer for AI coding assistants: clarifies intent, routes capabilities, reviews evidence, verifies results and writes back lessons.
- NANOSTACK BOOSTS CODE RELIABILITY
nanostack helps AI agents plan, check, test, and release more reliable code through a simple workflow wrapper.
- Normalize Similar CJK Glyphs
Folds visually similar CJK glyphs (Kangxi radicals and CJK radical supplements) into a canonical form, enabling PDF‑extracted text to match searchable Unicode.
- Bitemporal AI Memory System
Bitemporal memory system with six-variable decay, time-graded recall, TTL expiry, deduplication, and competitive retrieval using Neo4j, PostgreSQL, and Redis.
- SAFE DEPENDENCY INJECTION FOR SWIFT
SafeDI generates dependency injection code at compile time. Eliminates runtime errors and removes boilerplate. Supports testing via mocks without extra setup.
- Medical AI Dataset Provenance
Verified license and provenance records for medical AI datasets with information cards, JSON Schema, decision rules, and hash-based evidence index.
- SKILL SET FOR CLAUDE CODE AND CODEX
A collection of skills for interface design, reel editing, and improving AI agent workflow. Installed via CLI and invoked by commands in a session.
- Speaker for presentations
The Speaker project is a Codex skill for automating academic presentation preparation: it extracts text and structure from PPTX, performs OCR on images, generates speaker notes for slides, and writes the clean lecture text into PowerPoint comments.
- AGENT AUTHENTICATION AND PAYMENTS
Full-featured TypeScript chat agent with built-in authentication, credit system, and payments.
- GLYD: LOSSLESS AI COMPRESSION
Glyd reduces GPU memory by 33% without losing data, ensuring bit‑for‑bit identity. The project is written in Rust and supports CUDA, C‑ABI, and CLI for convenient benchmarking.
- PHP Session UAF Exploit Framework
Exploits CVE-2026-109618/619 PHP session.upload_progress use-after-free vulnerability to achieve remote code execution.
- Local Knowledge Graph Builder
Turns notes, web pages, PDFs, and media into a personal knowledge graph running entirely on your machine.
- DeepSeek Harness Linux Packager
Provides ready‑to‑run Linux packages (AppImage, .deb, .tar.gz) for DeepSeek Harness desktop, supporting x86_64 and ARM64 via native CI builds.
- Multi-LoRA Serving Architecture
Routes and composes hundreds of LoRA adapters on a single base model for efficient serving.
- 🤖 GPT-IMAGE-2 PROMPT ENGINE
Complete guide to editing and generating images with AI. Master class in GPT-Image-2 prompts.
- SuperLog: observability and debugging
SuperLog is an open-source TypeScript observability tool that automatically detects and fixes errors in code. It integrates with CI/CD and provides real-time application monitoring.
- Self-Hosted AI Agent Platform
Open-source, self-hosted agent platform using ReAct and Harness patterns, with AI customer support as the reference application.
- Tufte VDQI Plugin
A plugin for AI agents that implements Edward Tufte's data visualization principles — clear, accurate, and noise-free.
- Skills for AI code agents
The BuilderIO/skills project provides ready-made JavaScript skills that let AI agents write, refactor, and test code. It easily integrates with LLM frameworks and is extensible via NPM modules.
- NOOBOT CHEAP AI AGENT
Self‑hosted workspace for AI agents with low cost: support for tool calling, MCP, multi‑model routing and sandboxed execution. Enables creating multi‑agent workflows and LLM‑generated 3D character animation rendered with Three.js.
- LOCAL-FIRST AI TRACE WORKBENCH
tracecrate — a local tool for inspecting AI agent traces. Supports Claude Code, Codex, and OTLP logs, lets you compare runs, and exports privacy-conscious reports with no data leaving your machine. No backend or API keys required.
- Nexus Agent Context
Nexus is a shared context plane where AI agents and humans collaborate, exchange knowledge, and grow together.
- Claude2API Claude.ai Gateway
A Go + Docker service providing a Claude.ai‑compatible API gateway, an account pool, and a web mirror. It supports OpenAI‑compatible endpoints, streaming output, multimedia, and key management.
- EVALFORGE — LLM EVALUATION IN CI
EvalForge is a tool for LLM evaluation in CI/CD. Provides evaluator-neutral evidence, baseline regression gates, and JSON, JUnit, SARIF reports. Integrates with GitHub Actions.
- AI AGENTS IN YOUR CHROME
Extension gives AI agents access to your Chrome with saved session. Agents work in background tabs via CDP without disturbing you. Compatible with Claude and other LLMs for browser automation.
- Beginner AI Learning Path
A step-by-step AI roadmap covering foundations to production with free resources and hands-on mini-projects for each layer.
- Local-First AI Desktop Assistant
Capri is a Windows AI companion running local or API models with Hugging Face support, voice, memory, and permission-based desktop automation tools.
- PERMIT0 — AI Agent Error Protection
Authorize AI-agent actions before execution. Deterministic policies, simple integration from day one.
- Ultra-Fast Qwen3-TTS
A library for instant speech synthesis based on Qwen3, ensuring time to first audio under 50 ms at a load of 10 requests per second.
- CODEX CONTROL CENTER
codex-command-center — a local desktop hub for managing Codex projects, engineering tasks, sessions, Git changes, skills, and MCP configuration. All data is gathered in one interface, allowing quick switching between tasks and preserving work context.
- DeepSeek Local Setup Guide
A Russian‑language guide that walks you through installing and running DeepSeek models (V4, R1) locally on PC, Mac, or server using tools like Ollama, llama.cpp, LM Studio, vLLM, Open WebUI, and more.
- China Market Data Skill
Python CLI providing A-share quotes, K-lines, tick data, research reports, fund flows, and macro data for Claude Code and AI agents.
- Local-First AI Agent Harness
Run any AI model locally with MCP support and natural-language CLI. Zero dependencies, written in Go.
- LLM Post-Training Framework
VIME is a framework for post-training large language models using vLLM, oriented toward scaling reinforcement learning.
- CC-FLEET: CLAUDE CODE WITHOUT ANTHROPIC
Run Claude Code workflows, agent teams and subagents on any model — DeepSeek, GLM, Kimi, Qwen or your Codex, without an Anthropic subscription.
- OpenSearch-VL: Multimodal Search
Open recipe for training deep multimodal search agents with data curation, visual tools, and RL.
- Stock Market Correlation Tests
Tests whether stock correlations predict market movements using 66 pre-registered experiments and a triple pendulum analogy.
- Full-Stack Interview Prep Platform
Open-source interview prep platform covering Java, Python, databases, algorithms, LLMs, AI Agents, and RAG with real questions mapped to companies and roles.
- Persona SDK for LLM Apps
Defines persona cards, loads configurations, and adapts prompts across LLM providers like OpenAI and Anthropic.
- Interruptible Icon Animation
A Flutter library that smoothly transforms vector icons and allows interrupting the animation at any moment. Supports custom paths and morph speed settings.
- OpenClaw: Secure MetaMask SDK
OpenClaw provides a TypeScript-oriented toolkit for secure interaction with MetaMask: connecting wallets, reading balances, signing messages, and sending transactions across all EVM networks without seed-phrase access.
- HERMINAL: TERMINAL WITH VIETNAMESE IME
Native terminal for macOS with Vietnamese input support and coding-agent observability. Works offline, Swift.
- Loom: delivery system
Loom is an open harness that turns Claude Code, Codex, OpenCode, and other agents into reproducible software delivery systems. It provides a universal interface for automating build, test, and deployment.
- Research workspace
Open-source workspace for searching academic articles, downloading PDFs, inline citation with document highlighting, reading mode with Q&A, and exporting citations in four formats.
- Pixel Art Skill for Claude Code
Teaches Claude Code to create pixel art, animated game sprites, and visual effects through guided prompts and techniques.
- AI Engineering Tutorial Hub
In-depth tutorials on LLMs, RAG, and AI agents with practical code examples.
- Unified Video Generation Framework
Unified multimodal framework for video generation based on diffusion priors. SIGGRAPH 2026 / TOG.
- MULTI-CLI - MULTI-PROFILE MODE
Run multiple AI coding tool accounts simultaneously, saving time on switching.
- AI AGENTS EXCHANGE MESSAGES
The project lets AI agents in different clients (Claude Code, Codex) exchange messages directly. Works locally, open source, no central coordinator.
- One-Command Offline Emergency Library
Downloads a full offline emergency library with local AI, Wikipedia, maps, and survival guides in one command. Runs entirely without internet.
- MUXI Infrastructure for AI
An open platform for deploying AI agents in production. Simple integration, scalability, and monitoring.
- Zetsu: Personal RAG
Zetsu is a personal knowledge search system for offensive security based on Retrieval-Augmented Generation. It lets you upload your materials, build a vector index, and get accurate answers with sources cited.
- POV AI Storytelling
Practical framework for creating first-person stories using AI visualization.
- Native IDE for AI Agents on Mac
Lightweight terminal (5 MB) on SwiftUI for running parallel Claude Code and Codex sessions in isolated branches. No Electron or telemetry.
- Deep Research Generator
Professional skill for quickly creating detailed research reports in 19 languages with one CLI command.
- Comprehensive AI Engineering Course
Free 173-lesson course covering data, ML, deep learning, and AI engineering with 166 runnable labs. Covers math, code, systems, and security under CC BY-NC 4.0.
- HyperFlow: 8-Step LoRA
HyperFlow implements an eight-step LoRA adaptation procedure for MiniMax-H3 in the modular diffusers pipeline, simplifying training and video generation.
- Windows EPT Hooking Framework
A C library enabling Extended Page Table based hooking on Windows 10 and 11, allowing interception of memory accesses without modifying code.
- Xbox 360 Games on Linux
Recompiles Xbox 360 titles to run natively on Linux. Provide your own game files; only launchers are distributed.
- GordenSuper PPT AI
The project generates high-quality presentations as images using GPT and converts them into fully editable PPTX files, simplifying professional slide creation.
- PowerPoint/HTML Toolkit
Python library that converts PowerPoint presentations to HTML and back, preserving exact formatting. Ideal for RAG pipelines, LLM agents, and automated slide editing.
- ROUTEVSCODE DYNAMIC AI SWITCHER
routeVSCODE — Zero‑reload switcher of AI models and local proxy for VSCode Copilot Chat via 9Router. Allows instantly switching GPT‑4, Claude, DeepSeek without reloading the editor, preserving the chat.
- Lossless memory for AI
The project provides long-term memory for personal AI without data loss: each line is saved with a timestamp and never summarized.
- Leaked AI System Prompts
Aggregates leaked system prompts from ChatGPT, Claude, Gemini, Grok, Perplexity, Cursor, Lovable, Replit, and more, giving researchers transparent access to model instructions.
- PRODUCT‑MODE FOR AI TEAMS
The project provides a CLAUDE.md/AGENTS.md template with principles of problem framing, scope, tradeoffs, outcome measurement for PM, eng team. It works as a drop‑in file in Claude Code, Cursor, or Codex, based on Karpathy’s observations.
- Claude Code Gateway Bridge
Connects Claude Code CLI to Google Gemini models via Anthropic Messages API with observability.
- Self-Hosted Observability Platform
Self-hosted observability stack for metrics, logs, traces, dashboards, and alerts. Stores data in PostgreSQL or ClickHouse, accepts OpenTelemetry.
- FIGureMirror: AI Graph Styling
An automated AI-based tool that in one click styles your graphs to match the figures in scientific publications.
- JOANIUM - YOUR SMART AI ASSISTANT
Joanium — a smart and friendly AI assistant built with JavaScript. It answers questions, helps with tasks, and adapts to your style.
- Conker Xbox Recompiled for PC
Native recompilation of Conker: Live & Reloaded (Xbox) for Windows/Linux, built from your disc, with widescreen up to 240 FPS, enhancements, and online play via room codes.
- Typed Decisions With Probabilities
Open-weight Python model that outputs typed decisions with probabilities in one forward pass. Apache-2.0 licensed weights and code.
- Photocraft Test Corpus
Provides sample images and metadata used to test the photocraft image processing library. Includes various formats and edge‑case files for unit and integration tests.