GitHub 트렌딩을 그대로 나열하지 않고, Claude Code · RAG · 로컬 AI · 에이전트 워크플로우 · 평가 도구 · AI 앱 빌더 관점에서 실제 빌더가 쓸 만한 오픈소스 스택을 다시 정리합니다.
AI 엔지니어링의 기본부터 실제 서비스 구축까지 다루는 실습 중심의 학습 자료다. 에이전트와 컴퓨터 비전 등 다양한 AI 분야를 포함한다.
Learn it. Build it. Ship it for others.
모든 소프트웨어에 어그리게이트를 만들 수 있는 CLI-어니띵.
"CLI-Anything: Making ALL Software Agent-Native" -- CLI-Hub: https://clianything.cc/
Karpathy가 던진 "리서치 조직 코드" 실험. train.py 한 파일을 에이전트가 5분짜리 실험 단위로 자율 수정·평가·반복 — 자는 동안 ~100회 실험이 돌아가고 모델이 점점 좋아짐. 사람은 Python 안 만지고 program.md만 편집한다는 게 진짜 포인트.
AI agents running research on single-GPU nanochat training automatically
Manus를 오픈소스로 따라가는 '범용 에이전트'. 웹·코드·데이터 작업을 한 인터페이스에서.
Open-source generalist AI agent for everyday tasks.
개인 트레이딩을 위한 AI 에이전트로, 알고리즘 트레이딩 및 백테스팅 기능을 제공한다. LLM 기반으로 시장 분석 및 거래 결정을 지원한다.
"Vibe-Trading: Your Personal Trading Agent"
Meta가 직접 미는 풀스택 에이전트 프레임워크. inference·메모리·세이프티·툴콜이 하나의 API로 묶여 있어요.
A unified API for building agentic applications — inference, memory, safety, and tool-use, batteries included.
에이전트가 실제 브라우저를 제어해서 웹 작업을 자동화하게 해주는 라이브러리. 폼 채우기·스크래핑·테스트 모두 가능.
Make websites accessible for AI agents — browser control for LLMs.
OpenAI Realtime API와 결합해 전화·통화 수준의 음성 에이전트를 만들 때 사실상 표준 프레임워크.
A powerful framework for building realtime voice AI agents.
Pydantic 만든 팀의 type-safe 에이전트 프레임워크. validation이 엄격해서 프로덕션 적합.
Agent Framework / shim to use Pydantic with LLMs.
Devin 같은 자율 SW 엔지니어를 누구나 자기 환경에서 띄울 수 있는 오픈 플랫폼.
OpenDevin — a platform for autonomous software engineers.
ARIS는 Claude Code와 같은 LLM 에이전트와 함께 작동하는 가벼운 마크다운 스킬입니다.
ARIS ⚔️ (Auto-Research-In-Sleep) — Lightweight Markdown-only skills for autonomous ML research: cross-model review loops, idea discovery, and experiment automation. No framework, no lock-in — works with Claude Code, Codex, OpenClaw, or any LLM agent.
역할 기반으로 여러 AI 에이전트가 협업하게 만드는 프레임워크. 'CEO·개발자·QA' 같은 페르소나가 한 작업을 분담.
Framework for orchestrating role-playing, autonomous AI agents.
허깅페이스가 만든 '코드로 사고하는' 에이전트 라이브러리. 단순함이 무기이고 onboarding이 빨라요.
A barebones library for agents that think in code.
고정된 LLM 에이전트를 위한 재사용 가능한 자연어 스킬을 훈련하는 텍스트 공간 최적화 도구이다. 궤적 기반 편집과 검증을 통해 에이전트의 스킬을 개선하고 배포 가능한 형태로 만든다.
SkillOpt is a text-space optimizer that trains reusable natural-language skills for frozen LLM agents through trajectory-driven edits, validation-gated updates, and deployable best_skill.md artifacts.
Claude Code, Codex, Cursor, Pi 등 다양한 AI 에이전트를 오케스트레이션하는 오픈소스 프레임워크이다. 에이전트 간 협업과 정책 적용, 샌드박싱을 통해 복합 AI 시스템을 구축할 수 있다.
Omnigent is an open-source AI agent framework and meta-harness: orchestrate Claude Code, Codex, Cursor, Pi, and custom agents — swap harnesses without rewriting, enforce policies and sandboxing, and collaborate in real time from any device.
동일 예산으로 더 높은 지능 밀도를 제공하는 토큰 효율적인 AI 에이전트이다. 적은 토큰으로도 복잡한 작업을 처리할 수 있어 비용 효율적인 AI 솔루션 개발에 기여한다.
OpenSquilla — Token-Efficient AI Agent with same budget, higher intelligence density
자연어를 코드로 변환해서 사용자 컴퓨터에서 직접 실행. 데이터 분석·자동화·시스템 작업에 강해요.
Natural language interface for computers — runs code locally.
PR-단위 작업에 특화된 터미널 페어 프로그래머. 개인 사이드 프로젝트에 강합니다.
AI pair programming with edit/commit integration.
PM·아키텍트·엔지니어·QA 페르소나가 자연어 요구사항으로부터 한 회사처럼 SW를 만드는 프레임워크.
The Multi-Agent Framework: First AI Software Company, Towards Natural Language Programming.
Claude Code Skill 패키지로 바로 설치할 수 있어서, 에이전트가 한 번에 GPT Image 2 호출하고 결과를 파일로 떨궈줍니다.
GPT Image 2 prompt gallery, image prompt library, agentic skill, and CLI for OpenAI image generation/editing.
LLM 기초부터 멀티 에이전트 시스템까지, Agentic AI 학습의 모든 것을 담은 로드맵이다. 엄선된 240개 이상의 자료와 실습 예제로 심층 학습이 가능하다.
A trilingual (繁中 / English / 简中) learning roadmap for agentic AI: from LLM basics to multi-agent systems, with 240+ curated resources and hands-on examples. 中文 AI agent 學習地圖。
OpenAI가 만든 학습용 가벼운 에이전트 오케스트레이션. handoff 패턴이 깔끔해서 패턴 학습용으로 좋아요.
Educational framework exploring ergonomic, lightweight multi-agent orchestration.
에이전트에게 장기 메모리를 주는 OS-스타일 프레임워크. 컨텍스트가 길어질 때 유용해요.
Letta (formerly MemGPT) — agents with memory and persistence.
LLM이 작업을 완료하도록하는 self-healing 하네스입니다.
Browser Harness | Self-healing harness that enables LLMs to complete any task.
웹 파싱 대신 픽셀 데이터를 직접 활용하는 새로운 RAG 검색 방식이다. 복잡한 UI나 멀티모달 정보에서 확장성 높은 검색을 구현할 때 유용하다.
The end of web parsing. The beginning of scalable pixel-native search. link: https://pixelrag.ai/
에이전트 스케일링 법칙을 연구하는 학술 출신 프레임워크. 역할 플레이 패턴이 잘 정리돼 있어요.
First and the best multi-agent framework. Finding the Scaling Law of Agents.
Serenity에서 영감을 받은 AI 에이전트 스킬로, 공급망 병목 현상과 관련된 주식 리서치를 수행한다. Claude Code, Codex 등과 연동되어 투자 연구를 자동화한다.
Serenity-inspired Agent Skill for supply-chain bottleneck stock research
프로젝트 스펙을 던지면 AI가 폴더 구조부터 코드까지 만들어주는 에이전트. 초기 코드베이스 빠르게 부트스트랩할 때.
Specify what you want it to build, the AI asks for clarification, and then builds it.
코드 거의 없이 프롬프트 조합만으로 멀티-LLM 에이전트 만들 때 출발점.
An easy way to build prompts and agents.
Google DeepMind의 과학 분야 에이전트 워크플로 가속화 도구다. 30개 이상의 과학 데이터베이스와 연동하여 에이전트의 추론 정확도와 토큰 효율을 극대화한다.
GDM Science Skills to speed up agentic scientific workflows with better grounding and higher token efficiency. Integrate insights from AlphaGenome, AFDB, UniProt and 30+ other databases and tools.
코드 위협 모델링, 스캐닝, 분류, 패치 등 보안 작업을 자동화하는 Skill 세트이다. 사용자 정의 가능한 자율 스캐닝 하네스를 제공하여 코드 보안 프로세스를 강화한다.
Skills for threat modeling, scanning, triage, patching, plus an autonomous scanning harness you can /customize
Browser-use보다 먼저 나온 웹 자동화 에이전트. Selenium 기반 액션 모델 학습 자료가 풍부.
Open-source Large Action Model framework for browser automation.
단순한 task list + 실행 + 평가 루프로 자율 에이전트의 본질을 가장 짧게 보여주는 코드.
AI-powered task management system with goal-oriented planning.
AI 에이전트를 위한 통합 실시간 검색 엔진 Skill이다. 다양한 소스의 최신 정보를 에이전트가 효율적으로 탐색하고 활용할 수 있도록 지원한다.
Unified real-time search engine skill for AI agents.
MS 리서치가 만든 다중 모델 오케스트레이터. HuggingFace 모델을 LLM이 도구로 호출.
JARVIS — connecting LLMs with ML community models.
Google Antigravity의 강력한 기능을 활용하여 AI 에이전트를 구축하기 위한 Python 라이브러리다. Gemini API를 포함하며, Google의 AI 인프라를 활용한 LLM 에이전트 개발 시 첫 번째 선택지이다.
A Python library for building AI agents that leverage the full power of Google Antigravity.
임베드 가능한 작은 개발자 에이전트. '내 앱 안에 코딩 도와주는 어시스턴트' 만들 때 출발점.
The first library to let you embed a developer agent in your own app.
LangChain 생태계에서 에이전트를 위한 모든 기능이 포함된 하네스이다. 에이전트 개발 및 배포를 위한 통합된 환경을 제공해 효율성을 높인다.
The batteries-included agent harness.
터미널에서 git 기반으로 페어 프로그래밍하는 CLI. Claude/GPT 모두 호환되고 자동 커밋이 깔끔해요.
aider is AI pair programming in your terminal
LLM에 입력되기 전 로그, RAG 청크 등 다양한 데이터를 60-95% 압축한다. 토큰 비용을 줄이면서도 동일한 결과를 얻는 데 효과적이다.
Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 60-95% fewer tokens, same answers. Library, proxy, MCP server.
LLM 입력 토큰을 60~95% 압축하여 비용을 줄여준다. RAG 청크, 로그, 파일 등 모든 입력에 적용 가능하며, 라이브러리, 프록시, MCP 서버로 활용된다.
Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 60-95% fewer tokens, same answers. Library, proxy, MCP server.
에이전트가 세션을 넘어 사용자 사실을 기억하게 해주는 메모리 레이어. 사용자별로 정보를 구조화해서 다음 대화에 자동 주입.
Universal memory layer for AI Agents
LLM을 사용한 A/H/US 주가 분석 시스템. 다양한 데이터 소스, 실시간 뉴스, LLM 의사결정 도구.
LLM 驱动的多市场股票智能分析系统:多源行情、实时新闻、决策看板与自动推送,支持零成本定时运行。 LLM-powered multi-market stock analysis system with multi-source market data, real-time news, decision dashboard, automated notifications, and cost-free scheduled runs.
Python 기반의 AI 에이전트 개발 프레임워크다. 자율 에이전트 및 AI 어시스턴트 구축에 필요한 핵심 기능을 제공한다.
Agent Zero AI framework
LLM을 이용한 다중 에이전트 트레이딩 프레임워크.
TradingAgents: Multi-Agents LLM Financial Trading Framework
여러 에이전트가 대화하면서 협업하는 패턴을 코드로 만들 수 있는 마이크로소프트의 프레임워크. 코딩·연구·자동화에 적합.
A programming framework for agentic AI
AI 에이전트와 RAG(Retrieval Augmented Generation) 시스템을 위해 비정형 데이터를 구조화된 청크로 추출, 파싱하여 제공한다. 복잡한 데이터를 LLM이 쉽게 소비하도록 전처리하는 데 특화되어 있다.
Knowhere extracts, parses, and outputs structured chunks ready for AI Agents and RAG.
대형 모델 기반의 AI 어시스턴트입니다.
Open-source super AI assistant & Agent Harness. Plans tasks, runs tools and skills, self-evolves with memory and knowledge. Multi-model, multi-channel. Lightweight, extensible, one-line install. (formerly chatgpt-on-wechat)
에이전트가 세션을 넘어 사용자 정보를 기억하는 건 큰 장점이다. 다음 대화에 자동 주입하는 기능이 특히 좋다. 이 레포의 코드 구조가 간결해서 바로 쓸 만하다.