Real-Project Audit Results — by Category

minia2a · August 2026 · Project names withheld under responsible disclosure

We ran full AI audits on real, live smart-contract projects. Below, results grouped by project category — the vulnerability classes found and the report price. Names are withheld so no live project is exposed.

Audit results

Project type	Modules audited	Vulnerability classes found	Worst severity	Report price
AI-agent protocol	7 (bridge, lending, governance, router, task, timelock, validator)	8 classes — cross-chain double-claim, bad-debt, missing access control, tx.origin auth, no slippage protection, duplicate-signer consensus, fake-completion payout, self-destructible timelock	Critical	$1,000
Cash / escrow system	1 (core record-keeping)	5 classes — unauthorized init, insufficient-balance errors, global-variable pollution, flawed guard, fee miscalculation	High	$500
Escrow arrangement	1	4 classes — dead inventory check, unauthorized listing, state-machine bypass, unchecked transfer	High	$300

How pricing works

Report price scales with protocol complexity and modules audited. A single simple contract starts at $200; a multi-module protocol with bridge + governance + lending is $1,000. Every report includes exact locations and exploit mechanics.

Responsible disclosure. Vulnerabilities are reported privately to project owners. We do not publish details of live projects. Findings are for the project owner's use before an attacker finds them.

Is your project in one of these categories? Get an audit before an attacker does.
• Static Scan — $2 · AI Deep Audit — $20 · Full protocol report — from $200
AI audit is probabilistic — may miss vulnerabilities or report non-issues. Treat as guidance, not proof.