Trending AI Tools

The Caveman repo optimizes LLM prompts by reducing token count by 65% through primitive language generation.

MemPalace provides a highly optimized, open-source AI memory system for large language models.

The graphify repository enables the creation of queryable knowledge graphs from diverse code, data, and documentation sources.

The career-ops repo provides an AI-powered job search system with 14 skill modes and Go dashboard capabilities.

The open-design repo enables local-first, open-source design system generation and prototyping with 71 brand-grade templates.

OpenClaude is a cross-platform AI agent that runs on various systems and integrates with diverse tools.

The browser-harness repo enables self-healing browser automation for Large Language Models (LLMs) to complete any task.

This repository enables AI-generated HTML slide decks with Swiss layouts and image prompts.

Vibe-Trading enables users to backtest and execute algorithmic trading strategies using LLM-based decision-making.

The open-multi-agent repo automates task DAG generation from goals using MCP and live tracing.

The holaOS repository enables the creation of proactive AI work-streams through a runtime agent-harness.

The apfel repo enables on-device OpenAI-compatible AI capabilities on macOS via Apple Intelligence.

The hermes-web-ui repository provides a multi-platform AI chat dashboard for Hermes Agent management and analytics.

The garden-skills repo showcases a web design skills collection with AI-powered image generation capabilities.

DeepSeek-Reasonix provides a persistent AI coding agent for the terminal with prefix-cache stability.

This repository aggregates top open-source AI projects, models, tools, and infrastructure for machine learning and more.

The m_flow repo implements a bio-inspired Graph RAG memory engine for episodic and semantic memory.

This repository provides a fully localized and enhanced version of Superpowers with 6 original Chinese skills.

The claude-code-book repo provides a 420,000-word deep analysis of Claude Code's AI Agent architecture.

✨ The agentic HTML editor — your local AI agent writes the HTML, you ship it. 🚀 75 Skills × 9 Surfaces (magazine · deck · poster · XHS / tweet · prototype · data report · Hyperframes) 🛡️ Sandboxed preview · 📤 1-click to WeChat / X / Zhihu / HTML / PNG 🔑 Zero API key — Claude Code / Cursor / Codex / Gemini / Copilot / OpenCode / Qwen / Aider.

This repository enables private, offline, and airgap-ready on-device execution of Claude Code on Apple Silicon.

The anything-analyzer repo provides a comprehensive protocol analysis toolkit with AI-driven capabilities.

The agents-cli enables expert-level AI agent creation, evaluation, and deployment on Google Cloud via coding assistants.

The toprank repo provides open-source Claude Code skills for SEO, GEO, and Google/Meta Ads optimization.

A curated list of autonomous improvement loops, research agents, and autoresearch-style systems inspired by Karpathy's autoresearch.

OpenKB provides a scalable knowledge base for large language models (LLMs) with retrieval and generation capabilities.

The parlor repo enables real-time multimodal AI conversations on-device with natural voice and vision interactions.

This repository provides a comprehensive cheat sheet for AI engineering interview questions and answers.

PrismerCloud enables self-hosted, decentralized AI knowledge management through a knowledge-base and LLM tools integration.

开源微信 Bot 管理平台 + App 应用市场 | Self-hosted WeChat Bot Platform with App Marketplace | Lark · Slack · Discord · DingTalk · GitHub · Notion · 20+ Apps | AI Tools | 7 Language SDKs

This repository enables running high-quality, uncensored LLMs directly from a USB drive or SSD on any platform.

面向商业分析师的智能数据分析体。Intelligent Data Analysis Agent for Business Analysts.

Cross-CLI skill for Obsidian. Turns your vault into a living AI-first second brain across Claude Code, Codex CLI, Gemini CLI, and OpenCode. 32 commands, vault-first research, scheduled agents, write-time AI-first validator.

The little-coder repo enables efficient coding with smaller LLMs through optimized agent capabilities.

The turbovec repo provides a high-performance vector index for nearest-neighbor search.

Independently authored prompt templates for AI coding agents — system prompts, tool prompts, agent delegation, memory management, and multi-agent coordination. Informed by studying Claude Code.

The ThinkWatch repository provides a unified proxy for secure AI API access and MCP management.

One sentence creates an AI-driven world — generate maps, characters, and watch stories emerge on their own. 一句话生成一个AI自主驱动的世界.

PokeClaw enables on-device AI control of Android phones using a Gemma 4 microcontroller locally.

The openyak repo provides a local-first AI agent for desktop work with Ollama/Rapid-MLX support.

More is Different. A multi-agent world engine where AI agents live, talk, compete, ally.

Multi-model DAG-driven parallel AI film generation — parallel speedup scales with scene independence; Generate film scenes simultaneously instead of one by one; "把影视生成的执行图从拓扑序变成关键路径最优调度" ; 唯一把场景叙事依赖建模为 DAG、以 CPM 算法驱动并行调度的影视生成引擎

This repository enables local AI-driven knowledge graphing in Obsidian using Ollama and Karpathy's LLM Wiki pipeline.

Turboquant enables lossless KV cache compression for LLM inference with 7x longer context in pure C.

RDNA-native LLM inference engine in Rust.

Pure Rust Inference Engine

Project_Chronos accelerates zero-stall MoE inference with lookahead prediction and async DMA prefetching.

Drop-in prompt compression for production LLM apps. Cut your token bill 40-60% without changing your code. Python SDK, LLMLingua-2, MIT.

One-click Qwen3.6-27B inference on Windows. 158 tok/s on RTX 5090, 72 tok/s on RTX 3090. Native, no WSL, no Docker, no telemetry.

Self-hosted auto clustering AI agent OS for low cost consumer hardware like the computer you have, an Orange or Raspberry Pi or a Mac etc. Desktop shell, app store, agent deployment, distributed compute cluster. Memory by taOSmd.

SpectralQuant: Calibrated Eigenbasis Rotation and Water-Filled Bit Allocation for KV-Cache Compression

mlx-flash enables weight streaming for MLX, running massive models larger than RAM on Apple Silicon.