# Run Claude Managed Agents with Vercel Sandbox

DevFeed: [Run Claude Managed Agents with Vercel Sandbox](<https://devfeed.tech/articles/run-claude-managed-agents-with-vercel-sandbox-1074.md>)

Original publisher: [Read original article](<https://vercel.com/changelog/run-claude-managed-agents-with-vercel-sandbox>)

Author: Allen Zhou

Published: 2026-05-18T23:01:00Z

Content type: release

Language: en

Sources: [Vercel News](<https://devfeed.tech/sources/vercel-news.md>)

Topics: [Claude](<https://devfeed.tech/topics/claude.md>), [Vercel](<https://devfeed.tech/topics/vercel.md>), [Firecracker](<https://devfeed.tech/topics/firecracker.md>), [Security](<https://devfeed.tech/topics/security.md>), [Cloud](<https://devfeed.tech/topics/cloud.md>)

Tags: [agents](<https://devfeed.tech/tags/agents.md>), [claude](<https://devfeed.tech/tags/claude.md>), [cloud](<https://devfeed.tech/tags/cloud.md>), [firecracker](<https://devfeed.tech/tags/firecracker.md>), [firewall](<https://devfeed.tech/tags/firewall.md>), [low-latency](<https://devfeed.tech/tags/low-latency.md>), [sandbox](<https://devfeed.tech/tags/sandbox.md>), [security](<https://devfeed.tech/tags/security.md>), [vercel](<https://devfeed.tech/tags/vercel.md>)

## AI overview

Vercel Sandbox now supports Claude Managed Agents, combining Anthropic's managed agent loop with self-hosted execution on Vercel infrastructure. Each session runs in an isolated Firecracker microVM with firewall-based credential brokering, deny-by-default egress, and low-latency access to private networks and cloud workloads.

## Source excerpt

You can now run Claude Managed Agents with Vercel Sandbox. Claude Managed Agents handles the model, harness, tools, and session state. Self-hosting lets you bring the execution environment, so an agent's tool calls run on your existing Vercel infrastructure with your private APIs, internal services, and customer data. Each agent session runs in its own isolated Firecracker microVM, using the same infrastructure that powers 1B+ Vercel deployments with enterprise-grade security, availability, and performance. What you get: A managed agent loop from Anthropic, including the model, harness, tools, and session state A Vercel function control plane that spawns a sandbox per session Firecracker microVM isolation per session, with millisecond startup Credential brokering at the firewall, so secrets never enter the sandbox Deny-by-default egress with a domain allowlist Low-latency connectivity to your private network and cloud workloads Follow the step-by-step guide for a reference implementation and learn more in the Vercel Sandbox docs. Read more