Congratulations on the release
SwiftSync and proofless IBD are impressive milestones.
I wanted to ask a question about a use case we are currently exploring: running Utreexo inside a Trusted Execution Environment.
We are building an autonomous Bitcoin wallet architecture (PRECOP) where an AWS Nitro Enclave runs a Simplicity Bit Machine to enforce spending constraints before producing BIP-341 Schnorr signatures. Because TEEs have very tight constraints (~256MB RAM, no persistent storage, stateless restarts), running a full node inside the enclave is not practical.
Today the enclave receives PSBTs from an external bridge process and recomputes the sighash internally after overriding witness_utxo. However the enclave still cannot independently verify that the input UTXOs actually exist in the current UTXO set.
Utreexo seems like a clean solution here. The idea would be:
-
The enclave keeps only the Utreexo forest roots in memory (a few KB, updated with each block header)
-
The bridge provides inclusion proofs for every PSBT input
-
The enclave verifies the proofs against its local roots before running the Simplicity policy engine and signing
This would allow the enclave to verify UTXO existence with something close to full-node assurance while staying within a small memory footprint.
My question: is the rustreexo accumulator library stable enough to embed as a crate in a constrained environment?
Specifically we would only need the verify and update roots paths (not proving or deletion). Are those components well-tested independently of the full utreexod / Floresta node implementations?
Utreexo looks like a very natural primitive for trust-minimized signing environments.