agent-sandbox MCP Server
agent-sandbox/agent-sandbox
[](https://agentrank-ai.com/tool/agent-sandbox--agent-sandbox/?utm_source=badge&utm_medium=readme&utm_campaign=agentrank_badge) Agent-Sandbox is an easy-to-use, enterprise-grade sandbox platform for AI Agents — letting them securely run untrusted LLM-generated code, Browser use, Computer use, and deploy websites, among other tasks.
claude mcp add agentrank -- npx -y agentrank-mcp-server Overview
agent-sandbox/agent-sandbox is a Go tool licensed under Apache-2.0. Agent-Sandbox is an easy-to-use, enterprise-grade sandbox platform for AI Agents — letting them securely run untrusted LLM-generated code, Browser use, Computer use, and deploy websites, among other tasks. Topics: agent-sandbox, ai-sandbox, browser-use, code-executor, computer-use, sandbox, agent, container, mcp, ai-infra, kubernetes-native, e2b-compatible.
Ranked #4350 out of 24846 indexed tools.
Ecosystem
Score Breakdown
220 stars → growing community
Last commit 1mo ago → moderate activity
3/7 issues closed → many open issues
6 contributors → small team
No dependents → no downstream usage
Weights: Freshness 20% · Issue Health 20% · Dependents 22% · Stars 10% · Contributors 8% · How we score →
How to Improve
Badge all embed codes →
[](https://agentrank-ai.com/tool/agent-sandbox--agent-sandbox/?utm_source=badge&utm_medium=readme&utm_campaign=agentrank_badge) <a href="https://agentrank-ai.com/tool/agent-sandbox--agent-sandbox/?utm_source=badge&utm_medium=readme&utm_campaign=agentrank_badge"><img src="https://agentrank-ai.com/api/v1/badge/agent-sandbox--agent-sandbox" alt="AgentRank"></a> Matched Queries
From the README
<div align="center">
<h3>One Deployment. Every sandbox your agent needs.</h3>
<p>
<a href="https://agent-sandbox.github.io"></a>
</p>
<p>
Self-hosted, lightweight, and easy-to-use sandbox runtime for AI Agents<br/>
</p>
</div>
---
# Why Agent-Sandbox?
When you're building AI Agents, one of the hardest infra problems is running untrusted, LLM-generated code and actions safely — with **multi-session and multi-tenant** isolation, so one agent's runaway task never touches another's.
Each sandbox needs to be isolated on a **per-agent, or even per-user** basis, and needs to persist state across turns so an agent can keep context across multiple interactions.
[kubernetes-sigs/agent-sandbox](https://github.com/kubernetes-sigs/agent-sandbox) solves this with [AIO Sandbox](https://github.com/agent-infra/sandbox) and Kubernetes, but it talks to Kubernetes directly — great if you're a platform team, painful if you're an agent or a developer who just want Read full README on GitHub → Get the weekly AgentRank digest
Top movers, new tools, ecosystem insights — straight to your inbox.