0PricingLogin
AI Agents · Lesson

Sandbox Execution for Code Agents

Never run model-generated code on your host — use gVisor, Firecracker, or E2B sandboxes.

Why Sandbox?

Code agents (Devin, Claude Code, OpenDevin) run model-generated code. If you run that code on your host, the model can:

  • Read your secrets / API keys
  • Delete files
  • Make outbound network calls
  • Mine crypto

Sandbox = isolated environment that limits damage.

Levels of Isolation

LevelToolUse
Processsubprocess + chrootLow security, dev only
ContainerDockerDecent default
VM / microVMFirecracker, gVisorStrong isolation
Hosted sandboxE2B, DaytonaEasiest production

All lessons in this course

  1. Prompt Injection Defences
  2. Output Filtering (Llama Guard, NeMo)
  3. Sandbox Execution for Code Agents
  4. Access Control on Tools
← Back to AI Agents