Jordan

There are a couple of loosely-connected dimensions to this. Perhaps the most fundamental is that the idea of a **multichain wallet** is not well-defined. The Wallet Standard doesn't define this,...

@ekryski It's unclear if we really need anything like that. It can be left up to wallets for now and standardized later if needed, I suspect. That was added when...

I removed most of the TODO comments that have been addressed by changes or posed open questions that have been resolved. The few that are left should be resolved by...

Do we need anything else besides the chains package to close this?

Will reopen for now, feel free to close if no longer needed.

Looks good! Can you add a patch changeset?

Hmm actually, I need to think about this change and see how we're using the dependency. bs58 5 breakingly changes from Buffer to Uint8Array types.

Thanks for opening discussion on this. First a few unstructured thoughts about security that come to mind -- I haven't noodled on this very hard yet, but I could imagine...

> the threat of a dapp suggesting a malicious RPC A wallet that exposes a `signAndSendTransaction` (or `signTransaction`) API will typically use an RPC's `simulateTransaction` API. This works (in the...

> what if we optionally allow folks to include a restriction on the cluster name in their signed transaction This cannot be done without changing the wire transaction format. If...