About

OmniGuard Labs Research

We are a smart-contract security research lab. We do not flag vulnerabilities - we prove them, by executing the exploit on a real EVM and shipping a signed, independently replayable receipt.

Our public output is the Proof Wall, Incidents, the Classes catalogue, and the research behind them.

Why anonymous

We publish under an institutional byline - OmniGuard Labs Research - without individual names. In security research, the artifact is the credential: every claim we publish resolves to a signed receipt anyone can verify. A signature you can check is a stronger trust primitive than a name you cannot. Our detection model and prover internals are deliberately withheld as trade secrets; the Verdict Standard states exactly what we do and do not claim.

What we publish

Execution-proven vulnerability classes; historical exploits re-executed end-to-end on a mainnet fork; signed, hash-chained receipts for every proven verdict; and a free proof-backed honeypot checker. We downgrade to unverified when we cannot read a risk source - behavior, not a heuristic score. Everything is dated and independently verifiable.

Contact

Research, disclosure, and press: admin@omniguardlabs.com. Postal contact: see the contact page.