Ask HN: The new wave of AI agent sandboxes?

In the last couple of months, several new solutions for sandboxing AI agents have launched (microVMs, WASM runtimes, browser isolation, hardened tool containers, etc.). Curious to hear from people using them in production. Are they working as advertised, or are there still major tradeoffs around security, cost, and performance?

Here's my list of sandboxing solutions launched in the last year alone: E2B, AIO Sandbox, Sandboxer, AgentSphere, Yolobox, Exe.dev, yolo-cage, SkillFS, ERA Jazzberry Computer, Vibekit, Daytona, Modal, Cognitora, YepCode, Run Compute, CLI Fence, Landrun, Sprites, pctx-sandbox, pctx Sandbox, Agent SDK, Lima-devbox, OpenServ, Browser Agent Playground, Flintlock Agent, Quickstart, Bouvet Sandbox, Arrakis, Cellmate (ceLLMate), AgentFence, Tasker, DenoSandbox, Capsule (WASM-based), Volant, Nono, NetFence

12 points | by ATechGuy 2 days ago

8 comments

  • allinonetools_ 6 hours ago
    From what I have seen, most of these work fine for basic isolation, but things get tricky when agents need real-world access or longer workflows. There is always a tradeoff between safety and usefulness, and it is still not fully solved yet.
  • QubridAI 2 days ago
    They work, but with tradeoffs. MicroVMs are secure but slower & costly. WASM is fast & cheap but limited. Ultimately, to date, there isn't a perfect solution. A majority of people employ a hybrid solution.
  • agcat 2 days ago
    I know there are too many of them, found this benchmarks that my partner did - https://github.com/nickaggarwal/sandbox-test/blob/main/FULL_.... He is planning to add a few more of them.
    [-]
    • ATechGuy 1 day ago
      This is very useful. Thanks for sharing!
  • ex-aws-dude 2 days ago
    A tool so good that it requires 37 different sandboxes
  • Mooshux 10 hours ago
    [dead]
  • aimarketintel 1 day ago
    [flagged]
  • olivercoleai 2 days ago
    [dead]
  • cochinescu 1 day ago
    [dead]