Cashu mints will be massive, when combined with ARKs that can verify proof of reserves
Mint operator will have less power over mints that operate on top of hardware with trusted enclaves storing keys, check recent post by @calle about it
What if we added FROST?
It transforms the single-enclave model into a threshold scheme where multiple enclaves (or even a mix of enclaves and offline devices) jointly control the mint, eliminating the single-point-of-failure problem.
We can verify that given mint is using trusted enclaves, in fact this is how it could look like in BAO:
Attack vectors?
BAO would present it like this to users:
• Fake signature?
No — the hardware manufacturer's signature is the root of trust.
• Fake meaning?
Yes — if the code inside the enclave is backdoored or the verifier doesn't check MRENCLAVE/freshness.
• Advertised how?
Via a public .well-known/attestation endpoint carrying a manufacturer-signed quote.
• BAO's job?
Curate expected MRENCLAVE/PCR values from audited source code, verify quotes in CashuMintChecker, and reject stale or mismatched attestations.
"When Lambos?" was a meme, "when enclaves" is the real question to ask
#whenenclaves
Attack vectors?
BAO would present it like this to users:
• Fake signature?
No — the hardware manufacturer's signature is the root of trust.
• Fake meaning?
Yes — if the code inside the enclave is backdoored or the verifier doesn't check MRENCLAVE/freshness.
• Advertised how?
Via a public .well-known/attestation endpoint carrying a manufacturer-signed quote.
• BAO's job?
Curate expected MRENCLAVE/PCR values from audited source code, verify quotes in CashuMintChecker, and reject stale or mismatched attestations.
"When Lambos?" was a meme, "when enclaves" is the real question to ask
#whenenclaves
#graphify
