AIエージェント
Rakazo
Rakazo — an open-source, self-hosted AI agent for persistent teammates: bring your own model and computer, run it on web, desktop, iPhone and Android
★ 4.1k スターApache-2.0
ガイドを見る →open-multi-agentは、open-multi-agentが公開しているオープンソースのAIエージェントプロジェクトです（GitHub: open-multi-agent/open-multi-agent）。GitHubスターは6,990、フォークは2,437で、MITライセンスの下で公開されています。GitHubでの最終更新は2026年10月8日です。
Open Multi Agent — a self-hosted TypeScript agent runtime with durable approvals and verifiable run records.
Decide your governance floor upfront: which actions need human approval, what the egress policy allows, and which providers' keys the runtime may use.
Clone open-multi-agent/open-multi-agent from GitHub (or use the install path in the run steps below) and skim its README for the parts relevant to your goal.
Add @open-multi-agent/core to your TypeScript project from npm and configure your provider adapters.
Define one agent and one task, run it, then inspect the verifiable run record and test that an approval-gated step really pauses.
Approval gates survive restarts — a run can pause for a human decision and resume later, rather than living only in memory.
Every run produces a verifiable record, so what agents did can be audited after the fact.
A baseline of governance controls that child policies can only tighten, with enforcement that fails closed when a transport cannot be enforced.
Built-in adapters for major hosted providers and Chinese cloud models, plus local runtimes (Ollama, vLLM, llama-server) through baseURL and any OpenAI-compatible endpoint.
Set offline or allowlist egress, checked before a built-in adapter connects — your keys, your models, your network rules.
No telemetry, no OMA backend or account, and no analytics, license or update requests — it is a library you host.
Get the source from the official GitHub repository for open-multi-agent.
git clone https://github.com/open-multi-agent/open-multi-agent.git
cd open-multi-agentAdd the core package to your project exactly as the README's get-started section shows.
npm install @open-multi-agent/coreFollow the README get-started example to define an agent/team and run your first task with its journal enabled.
# Use the README get-started code (runAgent / runTeam) in your project — see READMEThe open-multi-agent project, through the open-multi-agent/open-multi-agent repository, with the site at open-multi-agent.com.
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.
You supply the model access — provider API keys for hosted models, or local runtimes such as Ollama for free local inference; the runtime itself is free and MIT licensed.
Typical frameworks optimise for getting agents running quickly. OMA optimises for governing them — durable approvals, verifiable journals, egress policy and zero telemetry — which matters once agents act in production under real accountability.
Clone the repository (or use the documented installer), follow the install and configure steps on this page, and run the smallest real task first — define one agent and one task, run it, then inspect the verifiable run record and test that an approval-gated step really pauses.
Rakazo — an open-source, self-hosted AI agent for persistent teammates: bring your own model and computer, run it on web, desktop, iPhone and Android
Autonomous AI agents for penetration testing — self-hosted, multi-agent, with a full web UI
Ultra-lightweight, open-source, self-hosted personal AI agent framework in Python with WebUI, tools, memory, MCP, multi-agent workflows, automation, and chat…
A portable AI agent runtime for the JVM — write one @Agent, run it on any Java AI stack.
Self-hosted AI pentesting agent that finds flaws and opens PRs to fix them.
Self-hosted ai agents: minimal ai coding agent (~1,000 lines of python) inspir.
LibreAIカタログに掲載されたAIエージェントプロジェクトの中で最もスターの多いもの — open-multi-agentが合わなかった場合に検討したい、現実的な候補リストです。
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.
カタログの中でopen-multi-agentとスター数が近いAIエージェントプロジェクト — 前後の棚もそのまま見て回れます。
A framework for building agentic apps
AppAgent — Tencent's multimodal agent that learns smartphone apps by exploration, then operates them.
Open-source 7B vision-language-action model and codebase for generalist robot manipulation.
A payments protocol for the internet — built on plain HTTP.
Product Management skills framework built on battle-tested methods for Claude Code, Cowork, Codex, and AI agents.
An embeddable, portable, branchable microVM that runs AI agents safely on your own machine.