awesome-ai-agents-2026/environments

Secure execution environments

This directory covers where agent-generated code runs safely — sandboxes, VMs, containers, and serverless runtimes that isolate untrusted code from your production infrastructure.

Why environments matter

When an agent generates and runs code, you're executing untrusted input. This is fundamentally the same security problem as running user-uploaded scripts, and it needs the same level of isolation. The question is always: what's the blast radius if this code does something malicious or buggy?

Isolation level	Blast radius	Speed	Example
Same process	Full system access	Instant	exec() in Python — never do this in production
Separate process	User-level access	Fast	subprocess with resource limits
Container	Namespace-isolated	~500ms cold start	Docker, Podman
microVM	Hardware-level isolation	~125-300ms cold start	Firecracker, E2B
V8 isolate	Memory-isolated	~0ms cold start	Cloudflare Workers

The right choice depends on your threat model. If the agent only generates pandas code, a container might be enough. If the agent writes arbitrary Python, you want a microVM.

Sections

Topic	File
Hosted sandboxes	sandboxes.md
Self-hosted VMs	vms.md
Container patterns	containers.md
Serverless runtimes	serverless.md

Decision guide

Start here: Who generates the code?
│
├── Your agent (untrusted code)
│   ├── Need <1s cold start? → E2B (Firecracker microVM)
│   ├── Need GPU? → Modal
│   ├── Need full dev environment? → Daytona
│   └── Self-hosting required?
│       ├── Drop-in Docker security? → gVisor (see vms.md)
│       ├── Full VM boundary at scale? → Firecracker (see vms.md)
│       └── K8s + VM isolation? → Kata Containers (see vms.md)
│
├── Your code (trusted, calling agent APIs)
│   ├── Short-lived functions (<15 min)? → Lambda or Cloud Run
│   ├── Need persistent containers? → Docker / Podman
│   └── Ultra-low latency, simple logic? → Cloudflare Workers
│
└── Mixed (trusted orchestrator + untrusted agent code)
    └── Run orchestrator on your infra, sandbox agent code in E2B or Modal