zkSecurity Finds Four Low-Severity Issues in Monero's Bulletproofs

Ten working days of review produced seven findings, and none of them was serious. zkSecurity audited the Generalized Bulletproofs implementation in monero-oxide, a Rust cryptography library, and reported no high or medium severity vulnerabilities, four low-severity issues and three informational notes. MAGIC Grants, the nonprofit research funder that contracted the work in June, published the result on 12 November. Monero confirmed it in a statement on X on 18 November. The result is one component, not the whole upgrade.
What the auditors checked
The reviewers were Mathias Hall-Andersen and Suneal Gong, named by both accounts this desk read and given no titles by either. Their work went past the usual bug hunt. zkSecurity wrote a new knowledge-soundness proof and formalised the protocol in Lean, a theorem prover that checks mathematical arguments by machine, covering completeness, zero-knowledge and soundness; the proofs sit publicly on GitHub. Crypto.news adds a boundary the other account leaves out. The formal work covered the interactive protocol in an abstract mathematical group, while the written audit covered the implementation. Those are different objects. Developers on monero-oxide answered all seven findings.
"It proves that the Generalized Bulletproofs protocol, in the form Monero implements it, is complete, zero-knowledge, and sound," the project said, quoting the findings published by MAGIC Grants.
Why 150 million outputs matter
Monero hides a spend today by mixing it with a small group of decoys, a construction called a ring signature. Full-Chain Membership Proofs, the change the project calls FCMP++, would replace that with a proof that the spent output belongs to the chain's entire set, which both accounts put at roughly 150 million existing outputs. The anonymity set stops being a handful and becomes the ledger. Zcash took a different route to the same goal, and froze funds in its oldest shielded pool when its NU7 upgrade went live.
The construction traces back to the Curve Trees paper by Matteo Campanelli, Hall-Andersen and Simon Holmgaard Kamp. Aaron Feickert, then at Cypher Stack, turned it into a technical description with security proofs. Brandon Goodell, also at Cypher Stack, later found problems in those proofs and proposed corrections. This audit sits at the end of that chain, not the start of it.
The parts still outstanding
A second audit, this one of the FCMP++ cryptographic implementation, went to Trail of Bits. MAGIC Grants announced it on 17 August, and neither account read reports a result. The second beta stressnet has been running since 6 May, when it started at block height 2,997,100. In October a Monero developer asked the community for 45 XMR to cover three months of work on core software, wallet improvements and privacy infrastructure. Scope is the recurring question in reviews like this, and Hacken made the same point about scope when it tallied third-quarter losses.
On the 500 XMR attached to an audit of FCMP++ integration, the two accounts split. Crypto.news says the Community Crowdfunding System had already funded it; CryptoDaily says the system set the sum aside. Funded and earmarked are different states, and neither account says which is right.
Neither account names a fork date
Nothing read gives a hard-fork date or an activation block height. CryptoDaily says further audits stand between the code and a release. The price has not answered the news either: XMR traded near $505 on Tuesday, little changed over 24 hours, down about 10% over seven days and about 16% over 30, on CryptoDaily's figures. Crypto.news carries a live ticker rather than a publication-time quote, alongside a header caption claiming a 12% jump toward $650 that its own article text contradicts, so that caption is not used here. Neither account is primary, and this desk did not read the MAGIC Grants announcement or Monero's post on X directly.
Read also: A Sixteen-Person AI Audit Returns 4,962 Findings Across 390 Bitcoin Repos