AgentPlaza
AgentPlaza seals each AI agent in its own microVM, holds it to laws you write and pays it from a…
FreeAICodeAI Agents1,026 impressions#79 of its week3 comments
Maker links to DevHunt
Comments
>log in to commentFor developers who need safe, governed AI agents, AgentPlaza isolates each agent in its own microVM, enforces custom laws, and manages payments from a central treasury. It lets you run a single agent or a fleet, monitor plans, arguments, and releases from one console. I’d love to hear what works and what could be improved 😊
Fatisco Stack Industries· 7d agoWhen an agent breaks one of your laws mid-task, does it just stop and wait for you, or can the orchestrator reroute the work to another agent?
It depends - you configure that behavior in the system. Plus there are law specifics that you set as either blocking , routing or just infractions.
More AI tools →74 free dev tools →
- Each agent runs in its own microVM
- Govern agents with custom laws (JHRL)
- Built‑in treasury for task payments
- Console for overseeing plans, conversations, and releases
- Supports fleet deployment across multiple nodes
- Open‑source, runs on Windows, macOS, Linux, Raspberry Pi
- Govern agents with custom laws
- Control budgets via a treasury
- Manage a fleet from a single console
- Open‑source for Windows, macOS, Linux, Pi
- Govern agents with custom laws you write
- Control budgets via an on‑chain‑style treasury
- Manage a fleet of nodes from a single console
- Supports Windows, macOS, Linux and Raspberry Pi
- Each agent runs in its own isolated microVM
- Treasury system to pay agents for model calls
- Run on a single machine or a fleet of nodes
- Web, desktop and CLI console for full oversight
- Open‑source, cross‑platform (Windows, macOS, Linux, Pi)
- Each agent runs in its own sealed microVM
- Tool calls are checked against customizable laws
- Budget‑controlled treasury for model usage
- Orchestrator plans work and delegates to agents
- Supports Claude, GPT, Gemini and local models
- Runs on Windows, macOS, Linux and Raspberry Pi
AgentPlaza runs each AI agent in an isolated microVM, enforces custom laws, and manages payment via a treasury.
- for
- Developers building and governing autonomous AI agents.
- pricing
- open source
- license
- Apache-2.0
Key features
- MicroVM isolation — Each agent runs in its own KVM-based microVM with snapshot and rollback support.
- Custom law enforcement — User-defined JHRL laws check every tool call and can block or record violations.
- Treasury & escrow — Tasks are escrowed in credits; completed work is paid and agents earn reputation.
- Orchestrator & council — A mayor-like orchestrator approves plans and a council votes on new agents and tools.
- Multi-engine support — Twelve LLM engines (Claude, GPT, Gemini, etc.) are available and governed by the same laws.
- Editor integrations — Agents can be driven from Zed, JetBrains IDEs, Neovim and return results as pull-requests.
Use cases
- Securely run a fleet of autonomous coding agents that must obey company policies
- Govern AI-driven data analysis pipelines with audit-ready transcripts
- Create sandboxed agents that access external APIs only through a controlled gateway
AgentPlaza vs alternatives
 AgentPlaza |  LangChain |  LlamaIndex |  Devin |  CrewAI | |
|---|---|---|---|---|---|
| Best for | MicroVM-isolated, law-governed AI agents | Framework for custom LLM applications | Data-centric LLM retrieval | Autonomous AI software engineer | Coordinating multiple AI agents |
| Pricing | Open source | Free | Free | Subscription | Subscription |
| DevHunt upvotes | 1 | 0 | 0 | 0 | 0 |
| Launched | Oct 2026 | — | — | — | — |
- AgentPlaza vs LangChain: LangChain is a library for building LLM apps, not a full orchestration platform with microVM isolation and law enforcement.
- AgentPlaza vs LlamaIndex: LlamaIndex focuses on data-grounded retrieval, while AgentPlaza provides sandboxed execution and payment management.
- AgentPlaza vs Devin: Devin offers autonomous AI software engineering but lacks built-in VM isolation and treasury features.
- AgentPlaza vs CrewAI: CrewAI enables multi-agent workflows but does not provide microVM sandboxing or custom law enforcement.
AgentPlaza FAQ
How are agents isolated?+
Agents run in KVM microVMs (Linux), Hypervisor.framework (Apple silicon), or Windows Hypervisor Platform, each with its own memory, disk and process space.
Can I change the enforcement rules at runtime?+
Yes, JHRL laws can be edited while the system is running and are applied to all subsequent tool calls.
Where are secrets stored?+
Model keys, git tokens and connector secrets stay on the hub and are accessed by agents only through the gateway.
What happens if an agent makes a disallowed call?+
The call is refused, the refusal is recorded on the agent’s record, and the orchestrator can intervene.
Can I run AgentPlaza on my own hardware?+
Yes, you can host the hub and nodes locally; all components are open source and run on Windows, macOS, Linux, or Raspberry Pi.
How do I integrate the agents with my IDE?+
Use the `agentplaza acp` command; editors like Zed, JetBrains IDEs and Neovim send tasks to agents and receive approvals as pull-requests.
Summarized by DevHunt from agentplaza.io · Oct 20, 2026. Details may change; check the official site.
What people say
“Built AgentPlaza: an open-source "town" for AI agents. Every agent lives in its own microVM: no API keys, no direct network.”
Reddit
Around the web
AgentPlaza: Your AI agents, governed.aiagentslisting.com
AgentPlaza: self-hosted AI agents, each in its own microVMagentplays.ai
Agent Plaza：给DSH 智能体增加零注册公开广场blog.yeyupiaoling.cn
Vibe CodingX
