Trending AI projects

REA is an MCP-based tool that enables AI agents to reverse-engineer software by inspecting binaries, applications, and runtime behavior to explain how features work.

A coding-agent skill that implements a structured, six-phase security audit workflow to systematically discover and verify vulnerabilities in a codebase.

A DSH plug‑in that adds free, no‑login LLMs (DeepSeek Flash, Kimi K3, etc.) to the model selector, provides a local OpenAI‑compatible endpoint, token‑saving features, aggregated provider support, combo routing, usage dashboard, and one‑click in‑app upgrades.

Hallmark is a design skill for coding agents that generates human-looking web pages from 48 themes, lints for AI tells, and supports redesign and audit.

VoiceStudio is a local-first desktop application for voice cloning, video dubbing, transcription, and audiobook creation that runs entirely on your own hardware.

Impeccable is a design-focused skill for AI coding agents that prevents generic, repetitive UI patterns by enforcing project-specific product context and running automated design audits.

DeepTutor is a personalized, agent-based learning platform that helps users organize knowledge, automate research, and interact with specialized AI tutors that retain long-term context.

A tool that removes ads, hot searches, and marketing notifications from Chinese Android phones by flipping existing settings switches and suspending ad packages — no root, no flashing, with per-device automation scripts and a desktop GUI.

An open-source agent harness by DeepSeek AI that uses a plugin-based architecture to simplify the creation and composition of AI agents.

Orca is a cross‑platform desktop IDE that orchestrates multiple code‑generation AI agents in isolated git worktrees, offering parallel execution, terminal splits, design‑mode UI capture, GitHub/Linear integration, remote SSH worktrees, and a mobile companion for monitoring and steering agents.

Ponytail is an open‑source prompt/plug‑in that equips LLM coding agents (Claude Code, Codex, Copilot, etc.) with a “senior dev” mindset: it reuses existing code, writes only the minimal needed implementation, and automatically reviews or audits changes for bugs, security, and missing tests. Install via a single plugin command or by copying a markdown skill file; then use slash commands like `/ponytail‑review` or `/ponytail‑audit`. Benchmarks in the README claim up to 53 % fewer lines, 45 % token savings, and 98 % of risky logic shipped with a test.

TradingAgents is an open‑source Python framework that orchestrates multiple LLM‑driven agents (analysts, researchers, trader, risk manager, portfolio manager) to evaluate market data and generate simulated trade decisions. It supports many LLM providers, point‑in‑time financial data, a CLI, Docker, and a Python API for back‑testing and research.

Graphify is a local-first knowledge graph generator that maps codebases and other files into a queryable graph, enabling AI coding assistants to explain, path-trace, and query instead of reading files. It parses code with tree-sitter for free, and supports docs, PDFs, and media via optional semantic passes.

Cindy is an open-source AI agent that runs locally on your own machine, combining multiple agent engines (harnesses like Claude Code and Codex), models, and tools into one agent that does real work in your projects. This repo contains the desktop and mobile client apps (Electron and Expo/React Native) plus shared packages, organized as a pnpm monorepo. The backend service lives in a separate repository. You can bring models via the official Cindy service, your existing Claude Code/Codex Coding Plan, your own API keys, or local models. The client is free and Apache-2.0 licensed, customizable through memory, skills, automation, MCP, and plugins, with official builds including opt-out TapDB analytics.

Elephant-Goldfish Model (EGM) is an open-source plugin for Spec-Driven Development with AI coding agents. It uses deep-context Elephant sessions to draft specifications and zero-context Goldfish subagents to stress-test them before code is written.

iFixAi is an independent auditing tool that tests AI agents in simulated business environments to uncover security risks, unauthorized actions, and operational misalignments.

DeepInfer is DeepSeek AI’s open‑source, high‑throughput inference engine for their V4.1‑Flash LLMs, offering optimized kernels, advanced parallelism, speculative decoding, deterministic sampling and structured‑output support on NVIDIA SM100 GPUs and Huawei Ascend NPUs.

ECC is an open‑source framework that adds a full engineering workflow (plan‑test‑implement‑review‑verify‑remember‑improve) to AI coding agents like Claude Code, Codex, and Kimi Code. It ships 71 agents, 311 skills, security scanning (AgentShield), and a guided installer (`npx ecc-universal@2.2.3 setup`). MIT‑licensed, with optional paid Pro features for private‑repo use.

OpenDuet is a self-hosted proactive AI agent that works autonomously, asks before acting, and keeps you informed via a morning brief and task triage. It's built on OpenBot and is in early development.

PPT Master is a free, open-source AI tool that converts documents (PDF, DOCX, web pages) into natively editable PowerPoint files. It runs as a workflow inside any agent-capable AI tool (Claude Code, Cursor, etc.), requires only Python 3.10+ to install, and produces real PowerPoint objects — slide masters, native shapes, data-backed charts — not flat images. It also supports template distillation, filling existing decks while preserving design, and adding transitions/animations. Key caveats: the AI model sets quality (recommended: Kimi K3 or Claude with ~1M context), results need polishing, and SmartArt is deliberately unsupported. Data stays local, no platform lock-in, and the only cost is your AI model usage. Built by a finance professional who wanted editable AI slides.

A plug‑in library that adds disciplined engineering workflows (spec writing, TDD, architecture review, etc.) to AI coding agents like Claude, Codex, or Copilot via slash‑command skills.

ThinkRail is an agentic IDE that brings an AI coding agent into isolated git worktree workspaces, with persistent project knowledge stored in living spec files. It offers a rich interface for parallel agent work, skills, MCP support, and a desktop or browser-based app.

Agent Reach is a CLI tool that enables AI agents to access and extract structured data from restricted, login-required, or video-based websites that standard search tools cannot reach.

magpie is a local gateway app that lets any AI coding agent (Claude Code, Codex, Gemini CLI, etc.) use any model provider (DeepSeek, Kimi, GLM, ChatGPT, etc.) from one interface. It edits agent configs, offers subscriptions across all agents, fails over between accounts when quotas run out, tracks usage and cost, supports plugins and routing groups, and runs on macOS, Windows, Linux, Docker, and Termux.

Claude SEO is a free, open-source SEO audit plugin for Claude Code. It automates full website audits in 10-15 minutes, checking technical SEO, content quality, schema, AI search readiness, local and e-commerce SEO, and delivers a prioritized action plan with clear verification steps.

HyperFrames is an open-source framework that turns HTML, CSS, and web animations into deterministic MP4 videos, specifically designed for AI coding agents to automate video production.

Strata is a free, MIT-licensed app that runs AI models locally on your own computer — no account or cloud needed. It offers a desktop app and a terminal app, supports several model families (Qwen and DeepSeek variants) with documented hardware requirements and speed benchmarks, and connects to coding tools like Claude Code and Codex CLI through OpenAI/Anthropic-compatible APIs. Works on NVIDIA, AMD, Intel Arc, Apple Silicon, and Ryzen AI Max hardware.

Turnitout is a programmatic, non-generative engine that removes AI watermarks, tracking metadata, and detection triggers from documents and source code while preserving original formatting and logic.

A private, policy-enforced runtime that lets AI agents safely read files, call APIs, and use credentials by confining each agent to an isolated sandbox and formally verifying any policy change before it goes live.

A coding agent that directs and renders short films entirely through code, offering 59 distinct visual styles and tools for users to create or customize their own cinematic projects.

FreeLLMAPI aggregates free tiers from 34 LLM providers into one OpenAI-compatible endpoint, offering smart routing, fallback, and per-key rate tracking, so you can use billions of free tokens without juggling multiple APIs.

A control system for a tiny biped robot that uses reinforcement learning policies to walk, roll, and interact with its environment.

LearnHarness is an open-source agent harness that teaches you through AI agents: it measures what you understand, plans a lesson step by step, and builds a persistent knowledge graph of your reasoning so learning carries across sessions.

XERJ is a local search engine for AI agents that indexes code, documents, and logs to provide precise, cited data retrieval, helping agents avoid token-heavy file reading and improve coding performance.

A curated collection of system prompts, persona definitions, and workspace policies designed to help researchers and developers study and configure the behavior of various frontier and open-weight AI models.

An agentic skills framework & software development methodology that works.

A collection of 76 game-dev skills (small reference files) plus a router that auto-selects the right ones based on your engine and task, letting AI coding assistants build games across 12 engines with version-pinned, validated knowledge.

Fractal is a distributed compute orchestration framework that uses Android phones as edge nodes to run federated learning and pipeline-parallel inference of large language models, preserving user privacy and device health.

Dossier is a self-hosted application for job seekers that keeps your career history in one profile, generates tailored resumes and cover letters for each job posting using only facts from your profile, and tracks every application from draft to offer. It runs on your own server, supports local AI models for privacy, and never sends anything without your confirmation.

AI Job Search is an open‑source, Claude‑Code‑driven workflow that automates the whole job‑application process: you set up a personal profile, scrape Danish (and other) job boards, have an LLM evaluate fit, draft LaTeX CV and cover letter, get a reviewer agent’s feedback, and output ready‑to‑send PDFs. It also includes interview prep, outcome tracking, Notion/Gmail sync, skill‑gap analysis, and custom template/portal extensions.

Archify is an AI-powered tool that converts descriptions or code repositories into interactive, shareable HTML diagrams for visualizing systems, workflows, and plans.

UI UX Pro Max is an open‑source AI skill that generates full design‑system specs (layout, style, colors, fonts, effects, checklist) from a natural‑language request, using 192 industry rules and searchable catalogs. It ships as an npm CLI (`ui-ux-pro-max-cli`) that integrates with many AI coding assistants (Claude, Cursor, Gemini, etc.). The basic version is free and open source; a premium tier adds brand‑identity and AI‑image generation.

A collection of agent skills that guide AI coding assistants to generate polished, high-quality web frontends with deliberate art direction and automated visual testing.

PhotoCraft is an open-source, Rust-based image editor that offers a familiar Photoshop-like experience, full PSD support, a large toolset, and agent-friendly automation, running natively on major platforms and in the browser.

OpenBot is a desktop app that lets you run and manage multiple AI agents (ChatGPT, Claude, Gemini, etc.) locally, giving each its own workspace, persistent chat, file handling, and inter‑agent messaging, with optional browser automation and a privacy‑first design.

Anatomy is a Claude skill that explains technical ideas by building them as interactive isometric figures (SVG, WebGL, or 3D). It runs in multiple AI coding agents, installs via npx/git/ZIP, and includes examples like a raptor engine and a dial indicator. It uses only Node built-ins and has no dependencies.

Postiz is an open‑source, self‑hostable social‑media management platform that lets you schedule posts across many networks and use AI agents (ChatGPT, Claude, etc.) for content creation. Built with Next.js, NestJS, Prisma, and Temporal, it can be deployed via Docker and offers a public API, SDKs, and automation integrations.

A free, MIT-licensed AI engineering curriculum with 542 hands-on lessons and 20 phases across Python, TypeScript, Rust, and Julia. Learners implement model internals, retrieval pipelines, and agent runtimes; test them; and keep reproducible evidence. Includes three scaffolded projects with local graders, study routes for Model Context Protocol and Agent Skills, certification tracks, and a web-based tutor that can run through a coding agent.