ИИ-агенты
SkillSpector
NVIDIA's security scanner for AI agent skills - check before you install
★ 19,7 тыс. звёздApache-2.0
Смотреть руководство →Voyager — проект с открытым исходным кодом из категории «ИИ-агенты» от MineDojo (MineDojo/Voyager на GitHub). У него 7 250 звёзд на GitHub и 706 форков, он распространяется по лицензии MIT и последний раз обновлялся на GitHub 3 апреля 2024.
Voyager — the open-ended embodied Minecraft agent that writes its own skills and keeps learning for life.
Set up the full environment first — Minecraft instance, Fabric mods, Node and Python parts — and decide whether to start lifelong learning fresh or resume from a checkpoint/skill library.
Clone MineDojo/Voyager from GitHub (or use the install path in the run steps below) and skim its README for the parts relevant to your goal.
Follow the README's four install parts in order: Python install, Node.js install, Minecraft instance install, and Fabric mods install.
Start lifelong learning, watch Voyager propose and attempt its first tasks, and inspect the skills it writes into its library.
An agent that continuously explores, learns new skills and takes on harder tasks without a fixed endpoint.
Learned skills are stored as code in a skill library and reused — you can also load a learned library to run specific tasks.
Voyager proposes its own next tasks based on its progress, rather than following a hand-written task list.
Environment feedback and execution errors feed back into prompting so the agent debugs its own attempts.
Learning can resume from a checkpoint, and a learned skill library can be pointed at a new specific task.
The official codebase for the Voyager paper, with the project site and arXiv paper linked from the repo.
Get the source from the official GitHub repository for Voyager.
git clone https://github.com/MineDojo/Voyager.git
cd VoyagerComplete the Python and Node.js installs from the README, including the package install and TypeScript build steps it lists.
pip install -e .
npm installWith the Minecraft instance and Fabric mods running and your login configured, start Voyager's lifelong learning from the README's getting-started code.
# Configure azure_login (recommended) in the getting-started snippet, then start lifelong learning — see READMEMineDojo, through the MineDojo/Voyager repository, with the project site at voyager.minedojo.org.
Yes — the project is open source under the MIT license as recorded for this listing. You can use, modify and distribute it under the MIT terms.
Voyager's reasoning is driven by an LLM, so you need the model access its setup requires (the README recommends Azure login for the Minecraft connection and its GPT-class model setup); the code is free under MIT.
Scripted bots (for example Mineflayer scripts) do exactly what you coded. Voyager writes, tests and stores its own skills under an automatic curriculum — a research system for open-ended learning, with a much heavier setup than a bot script.
Clone the repository (or use the documented installer), follow the install and configure steps on this page, and run the smallest real task first — start lifelong learning, watch Voyager propose and attempt its first tasks, and inspect the skills it writes into its library.
NVIDIA's security scanner for AI agent skills - check before you install
A CLI that writes and maintains agent-ready docs for your codebase.
An auto-company that works 24/7 on your own PC — 14 agent roles, a dashboard, real workflows.
Agno's self-learning data agent: grounded text-to-SQL answers over your database that improve with use.
Gini: a personal agent runtime that remembers, learns skills, and reaches you with the right control in time.
Run chops locally: your ai agent skills, finally organized. a macos app to.
Другие проекты из категории «ИИ-агенты» с наибольшим числом звёзд в каталоге LibreAI — реалистичный список альтернатив, если Voyager вам не подходит.
A personal AI that really does things — on any OS, with your data staying yours.
An agentic skills framework & software development methodology that works.
Agent harness optimization system with skills, memory and security for Claude Code, Codex and more
The agent that grows with you — a Python framework for adaptive AI agents.
DeepSeek's plugin-based agent harness where everything is a plugin.
Соседние проекты из категории «ИИ-агенты» в каталоге, ближайшие по числу звёзд к Voyager — продолжайте изучать полку по обе стороны от него.
Programmable guardrails to keep LLM systems safe and on rails.
The Enterprise-Grade Multi-Agent Orchestration Framework. Website: https://swarms.ai
OpenTelemetry-native observability for LLM apps.
Product Management skills framework built on battle-tested methods for Claude Code, Cowork, Codex, and AI agents.
Put guardrails on LLMs: validate outputs before they reach users.
Open-source 7B vision-language-action model and codebase for generalist robot manipulation.
Бесплатно и открыто — забирайте исходный код, ставьте звезду репозиторию и начинайте создавать.