[Feature Request] Musig2/MUSI2 Support

[OliverOffing]

Hey, could we add MuSig2 support to the wallet? It's a new multi-signature scheme that boosts privacy and cuts down transaction size. This would really enhance the wallet and attract users who need advanced crypto features. Thanks!

[Bicaru20]

Based on the BIP-327, I think the first steps of the implementation could look like this:

Option 1
One option I could think of based on BIP-373 is:

1. Step A (on wallet):

• Participants share their PubKey.
• The wallet computes the PubKey aggregation (KeyAgg).
• Then asks each party to sign a PSBT that contains all MuSig2 fields (key‑agg, nonce‑agg).

2. Step B (on SeedSigner):

• Scan one PSBT QR.
• Generate your Nonce (NonceGen)→ Scan back pubnonce into wallet → Sign partial → return PSBT.
• Wallet finishes aggregation and signature.

Option 2

1. Implement KeyAgg algorithm to calculate the aggregated public key of all participants. Prior to this, the participants must share their PubKey, so in the seedsigner environment, this would be done by scanning the QRs of each PubKey of the participants.
2. Implement NonceGen algorithm to compute the secnonce (secret nonce) and the pubnonce (public nonce, the one that will be shared). Here I'm thinking to put an option to allow the user to export the pubnonce in QR mode, so it can be shared with the others.
3. Implement NonceAgg to calculate the aggregate nonce. Here, we would need again to scan the pubnonce of the other participants in the group, otherwise we cannot calculate the aggregate nonce.
4. Now it is time to create the partial signature. Again, to export the signature, we can use a QR code.
5. Finally, to aggregate the signatures, we will need once again to scan the signatures of the other users.

I haven't mentioned, but in the last two points, we would also implement the verifications to verify the partial and aggregate signature.

The problem I have with this implementation is that it is not comfortable for the user, as it requires scanning QRs back and forth.

Would be nice some feedback on this to know if Option 2 is in the right direction, or if for the seedsigner it would be better a simpler implementation and require the help of external software (like sparrow in option 2) for some steps.

[kdmukai]

The SeedSigner model is inherently limited by / dependent upon additional outside wallet software.

So the first question is: which wallets support musig2 at the moment? Nunchuk has their musig2 + taptree solution (which strikes me as somewhat worryingly complex / ambitious / too early). And does it include other dependencies that would be blockers for us (e.g. taproot miniscript)?

And then what, if any, support does any wallet software have for interacting with a signing device for a musig2 flow? What communication standards exist, if any? And/Or is there QR-based support?

If there isn't a signing device flow, then would we be building this flow speculatively, hoping that a wallet project adds support for what we'll have built?

Broader question: my rough understanding of musig2 is that there are multiple rounds of interactivity required that are unavoidable. Is multi-round n-of-n even sensible for SeedSigner?

tldr: My sense is that it's too early. We depend on other parts of the ecosystem to adopt these new approaches first and afaik there's not much out there right now.