We attack your protocols and systems before someone else does.
BCTI is an Andorran security firm with a Web3 focus. We audit smart contracts with manual review, fuzzing and formal verification, attack the apps and keys around them, and run penetration tests and red teams on traditional infrastructure.
$ bcti audit --repo ./vault-protocol
[*] review 1,840 nSLOC · line by line
[*] fuzz 2.1M calls · 14 invariants
[!] broken deposit_mints_shares
↳ deposit(1)→donate(1e18)→deposit(5e17)[+] prove withdraw ≤ balance · all paths
[✓] report fixes verified
$
Web3 security
From line-by-line review to mathematical proof
Smart contracts are public, hold real value and are hard to patch. We stack three layers of assurance on your code and also test the Web 2.5 layer around it, where many of the largest losses actually begin.
Three layers of assurance
L1
Manual review
Auditors read every line, model how value moves through your protocol and hunt for the logic, economic and integration flaws no tool can see.
Best atDesign, logic and economic flaws
L2
Invariant fuzzing
Stateful fuzzers run millions of random transaction sequences against the properties your protocol must never break.
Best atEdge cases and broken invariants
L3
Formal verification
Symbolic execution and provers check critical properties for every possible input and state, not just the ones a fuzzer happened to try.
No surprises: you know what we will test, when, and what you will receive.
01
Scoping
We agree on objectives, scope and timelines, and sign the authorisation and NDA. For code we fix a commit hash; for systems, the targets and test windows.
02
Threat modelling
We map your attack surface the way an adversary would: contracts, privileged roles and dependencies, or exposed assets, technologies and people.
03
Testing
Manual review, fuzzing, formal methods and hands-on exploitation, with a working proof of concept for every serious finding.
04
Reporting
An executive summary plus a detailed technical report with evidence, severity ratings and prioritised remediation.
05
Fix review
Once you have fixed the issues, we review the fixes or retest your systems, and update the report.
Why BCTI
Findings that matter, explained so you can act
Aligned with
OWASP
OWASP SCSVS
EEA EthTrust
PTES
NIST SP 800-115
MITRE ATT&CK
Toolchain
Foundry
Echidna
Medusa
Halmos
Certora
Slither
Manual and expert-led
Tools are a starting point, not the deliverable. Every engagement is run hands-on by specialists who think like attackers.
Proof, not opinions
Every serious finding comes with a working proof of concept: an exploit test for contracts, reproducible steps for systems. That leaves no debate about impact.
Confidential by design
Signed NDA from day one and strict handling procedures for every piece of code and evidence we receive.
Standards and open tooling
Our work follows recognised standards, and the fuzzing and verification suites we write are built on open tools your team can keep running.
Manual review, invariant fuzzing and formal verification find different bugs. What each technique really proves, where it falls short and how to combine them.