cd /news/ai-agents/flashvmm-jail-your-ai-agent-in-a-mic… · home › topics › ai-agents › article
[ARTICLE · art-170008] src=flashvmm.com ↗ pub= topic=ai-agents verified=true sentiment=↑ positive

FlashVMM – jail your AI agent in a MicroVM that boots in 18 ms

FlashVMM launched a sandbox that gives each AI agent task its own FlashVM, a custom-built microVM the company claims boots warm in milliseconds and is discarded when the task ends. The tool installs via a PowerShell script, runs agents such as Claude unchanged inside a hypervisor boundary that sees only the project folder, and exposes the full VM lifecycle through flash-mcp, an MCP server offering 21 tools including create, fork, snapshot, restore and teardown. FlashVMM said warm pools and snapshots keep kernel boot off the hot path, and a GPU pool loads a model once on the host GPU for all permitted VMs, using Metal on Apple silicon and Vulkan on Linux and Windows.

read2 min views4 publishedNov 15, 2026
FlashVMM – jail your AI agent in a MicroVM that boots in 18 ms
Image: source

A sandbox for AI runs: every agent task gets its own FlashVM. It's claimed warm in milliseconds, runs whatever the agent wrote behind a real hypervisor boundary, and is thrown away when the task ends. One install, no account, no cloud. We built our own virtual machine for this: FlashVM, leaner and faster than a conventional microVM.

PowerShell as Administrator. It enables the hypervisor if needed, picks x64 or ARM64 and verifies the checksum.

irm https://flashvmm.com/install.ps1 | iex

Then jail the agent you already use, unchanged: flash agents setup, then cd my-project && claude. It runs in its own VM and sees only that folder. The full tour: flash demo.

What it's for

Let the agent run anything. Keep your machine.

Run code an agent wrote

Scripts, package installs, test suites, and Docker builds run inside a VM with its own kernel. Cap CPU, memory, disk and network only when you want to, and decide what it may reach with --egress. Nothing touches the host.

flash-mcp hands Claude, Cursor or any MCP client the whole lifecycle in 21 tools: create, run commands, copy files, fork, limit what it can reach, call APIs without holding the key, watch, snapshot, restore, tear down.

Warm pools keep spares already running, so a task claims one instead of booting. Snapshots restore a prepared environment in milliseconds, and flash fork branches a running VM into copies that carry on from where it was.

A GPU pool loads an AI model on the host's GPU once and every VM you allow sends it requests. No VM touches the GPU or needs a driver, and snapshots and warm pools keep working. Metal on Apple silicon, Vulkan on Linux and Windows.

The task takes a warm spare from the pool, or restores a snapshot. No kernel boot on the hot path.

2

Run

The agent's commands run in the guest with flash exec; exit code, stdout and stderr come back structured. Files go in and out with flash cp; logs, metrics and events stream back live.

3

Discard

The VM is removed when the task ends. Nothing it did carries into the next run, unless you snapshot it.

flash claim spare1 --entry 'npm test'   # a warm spare from the pool
flash logs spare1
flash rm spare1

Same CLI, same HTTP API and same MCP tools on every OS. Full walkthrough →

Give it a run.

No accounts, no dashboard. A CLI, a hypervisor, and an MCP server.

── more in #ai-agents 4 stories · sorted by recency
── more on @flashvmm 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/flashvmm-jail-your-a…] indexed:0 read:2min 2026-11-15 · —