Concept ACK, though a significant part of the sign/combine logic is being changed in #13425.

One possibility is only supporting these large multisig using PSBT, and not through the existing partial raw transactions interface. That should be very simple to write, but more restrictive.

Otherwise, the code for extracting existing signatures from a partial raw transaction into a key/sig map will need to be combined with this.

• BIP 174 PSBT Serializations and RPCs #13557 (BIP 174 PSBT Serializations and RPCs by achow101)
• Moving final scriptSig construction from CombineSignatures to ProduceSignature (PSBT signer logic) #13425 (Moving final scriptSig construction from CombineSignatures to ProduceSignature (PSBT signer logic) by achow101)

If you consider this pull request important, please also help to review the conflicting pull requests. Ideally, start with the one that should be merged first.

I'll rebase once #13557 is merged, making it PSBT-only.

Is there a reason for OP_NUMEQUAL rather than OP_EQUAL per BIP 19?

@luke-jr had no idea there was a BIP. I used a stack overflow answer. Will investigate.

@luke-jr Interesting! Reading that BIP, I believe the n-of-n template suggested there is incorrect. In case all signatures are valid, it will leave the stack empty, which is interpreted as failure.

Closing in favor of Future Work with descriptor language stuff