AI Audit Found Real Logic Bugs in a Live Escrow Contract

minia2a · August 2026 · AI Contract Audit · Static Scan

We ran our AI auditor against an active open-source escrow contract. It found high-severity logic flaws — real business-logic bugs a static regex scan would miss. Details are withheld under responsible disclosure.

Findings — High, under responsible disclosure

HIGHInventory check that can never fail

We found a stock-availability check that is logically always true — the "out of stock" guard is dead code, so any quantity (including zero) passes.

Full details withheld. Project owners: contact for the report.

HIGHUnauthorized listing

We found a listing function callable by anyone with no seller verification — fake items can be listed under arbitrary seller names.

HIGHState-machine bypass

We found settlement functions that don't verify the item is in the expected state, allowing the escrow flow to be finalized out of order.

MEDIUMUnchecked token transfer

We found a token-transfer return value ignored — balances update even if funds don't move.

Real escrow bugs found by AI in minutes — the kind that let an attacker disrupt or drain a live escrow. Details available to the project owner under responsible disclosure.

Full report available to project owners. Contact us for the complete vulnerability report.
AI Deep Audit — $200 (business-logic & economic flaws + responsible disclosure)
Static Scan — $2 (10 vulnerability patterns)
AI audit is probabilistic — may miss vulnerabilities or report non-issues. Treat as guidance, not proof.