User Guide

How to run an audit

Two ways to scan a contract on AuditSmart — a fast multi-agent Standard Audit, and an experimental Quantum Audit powered by IBM and AWS quantum circuits. This guide walks through both.

What is a Standard Audit?

A Standard Audit runs your contract through a 10-agent pipeline — 8 specialised Groq LLaMA agents, Google Gemini, and the Slither static analyzer — all in parallel, then merges and de-duplicates the findings. It completes in under 60 seconds and is the right starting point for almost every contract.

Step by step

1

Open the scanner

From the dashboard, go to Scan Contract (or click New Audit). This is your entry point for every standard audit.

Open Scanner
2

Paste your contract

Paste your full Solidity source — include all relevant contracts and imports, not just the vulnerable function. Pick the target chain and, if relevant, the token standard (ERC-20, ERC-721, ERC-4626…). The Solidity version is auto-detected from your pragma.

3

Run the audit

Hit Run Audit. Our pipeline fans out across 10 agents in parallel — 8 specialised Groq LLaMA agents (reentrancy, overflow, access control, logic, gas/DoS, DeFi, backdoor, signature), Google Gemini, and the Slither static analyzer — then de-duplicates overlapping findings.

4

Review the findings

Results arrive in under 60 seconds: an overall score, a severity breakdown, and each finding with a title, description, the exact line number, the offending snippet, and a concrete fix recommendation.

5

Export or share

Download a branded PDF report for your records or auditors, or share a public link to the result.

View History

Reading the findings

Every finding is tagged with a severity. Triage from the top down — clear Critical and High issues before anything else.

Critical

Funds at direct risk. Do not deploy — exploitable, high-impact (e.g. reentrancy on withdraw, missing access control on mint).

High

Serious flaw that can be exploited under realistic conditions. Fix before mainnet.

Medium

Meaningful issue with limited impact or harder preconditions. Should be addressed.

Low

Minor concern or hardening opportunity. Safe to deploy but worth fixing.

Info

Best-practice notes, gas hints, and style suggestions. No security impact.

Tips for the best results

  • Paste the complete contract, including imported libraries and interfaces — context improves detection.
  • Make sure your code compiles (in Remix or Hardhat) before submitting to avoid compilation errors.
  • A zero-findings result is not a guarantee of safety — pair high-value contracts with a manual review.
  • Need exploit PoCs and production-ready patches? Use Deep Audit (Claude Opus).