proposal-shadowrealm VS LavaMoat

ShadowRealm: ECMAScript Proposal, specs, and reference implementation for Realms [Stage 3 -> 2].

You can't out of the box. There is an open issue on the Node.js repositoryto let the node:vm module to use the vm's context, but it is still open. It seems that the Node.js core team is interested in fixing this problem by implementing the new ShadowRealm spec, and I think we will make some progress during 2023.

ShadowRealms — a successor of the Realms proposals, this API is intended for use cases exactly like plugins or extension systems, providing an option for creating distinct global environments to run the code in. While not entirely secure on its own, this API could provide a strong foundation to build actual extension systems on the Web. That said, 4 years later, the TC39 proposal is currently only at stage 3, not implemented by any browser;

With --no-isolate it was 2.8x faster than vitest and 1.7x faster than Jest, but 19 tests failed (see table above). Some people report issues with watch mode when using --no-isolate. So I decided to not pursue it any further. Once the vm module that Vitest relies on supports ESM, or when the amazingly named Shadow Realms are added to JavaScript, we will likely get this performance boost for free without the downsides.

If ShadowRealms are ever added to EcmaScript (and implemented into V8/Node) they'll allow for a different approach to isolating code that would be faster without the downsides of sharing global.

Eventually I will get around to creating a "proper" package by just grabbing all the JS that is loaded by the code in the iframe, and bundling it up. We really need the ShadowRealm proposal to go through because the perchance engine messes with a lot of JS internals, so it would mess up the rest of your app. Could do it in a WebWorker, but then everything would have to be async.

The upcoming JavaScript Shadow Realms proposal looks like it solves a similar problem: https://github.com/tc39/proposal-shadowrealm/blob/main/expla...

Browser-based JavaScript doesn't yet have a way to isolate code fully in this manner though there is a new JavaScript feature on the way that would provide this capability. Its called ShadowRealm and would basically give you a new global context to execute code that's completely separate from your main document code.

Just yesterday I watched a talk [0] at WarsawJS about LavaMoat [1], a set of tools to protect against malicious behaviour from npm dependencies. Guess it’s time to look into it deeper.

[0]: https://naugtur.pl/pres3/lava/2023end.html

[1]: https://github.com/LavaMoat/LavaMoat

You are basically talking about Lavamoat. It provides tooling and policies for SES, which aims to make it into standards.

https://github.com/LavaMoat/LavaMoat

Enter: lavamoat. https://github.com/LavaMoat/LavaMoat

You should check out Lavamoat: https://github.com/LavaMoat/LavaMoat

It attempts to do what you're essentially describing. It was built by the MetaMask team, where supply chain attacks are an obviously huge risk.

I've spent some time trying to get it working in an app, but haven't been able to get it all the way working. It's still pretty beta and not well documented.

And it is happening right now. Github is opening the GitHub Advisory Database to community submissions. Awesome community NodeSecure builds cool things like scanner and js-x-ray. There are also lockfile-lint, LavaMoat, Jfrog-npm-tools (and I am sure there is more).

There is a proposal to add OCAPs on a language level in TC39[0]. There is already a drop-in implementation which already works in both Nodejs and browsers[1].

As a developer who wants to sandbox your own (recursive) dependencies, this is made accessible today in Lavamoat[2]. Basically a package or app can provide a policy manifest specifying which capabilities (e.g. network or filesystem access) should be granted for each dependency. Also comes with a tool that will auto-generate a starting point from your existing dependency tree.

IMO this is the future. Currently it does come with a performance penalty but hopefully this idea will catch on and make it into runtime implementations.

Lavamoat is still marked as "preprod" on npm but talking to the author it's a matter of days or weeks until the first stable release.

[0]: https://news.ycombinator.com/item?id=30703817

[1]: https://github.com/endojs/endo/tree/master/packages/ses

[2]: https://github.com/LavaMoat/LavaMoat