A smart contract audit checklist is the preparation work your team finishes before an external auditor reads a single line: a frozen commit, a written threat model, high test coverage with fuzzing, clean static-analysis output and clear documentation. Teams that arrive prepared get deeper findings, fewer review rounds and a launch date that holds.
Why audit readiness matters more than the audit itself
Auditors are paid for time. If they spend the first week working out what your protocol is supposed to do, chasing a moving codebase or reporting issues your own linter would have caught, that is time not spent on the subtle economic and logic flaws that actually drain funds. A prepared codebase turns the engagement into a focused review of genuine risk.
Readiness also changes the outcome after the report. When findings come back against a stable, well-tested codebase, fixes are small and re-verification is quick. When they come back against code that was still changing, the fix review can become a second audit.
If you already have contracts written and want a senior engineer to tell you how far they are from audit-ready, send us your repository details and timeline and we will come back with a clear readiness plan.
The smart contract audit checklist, step by step
1. Freeze scope and the commit hash
List every contract in scope, its line count and its dependencies. Tag a single commit and agree that nothing in scope changes during the review. Note what is explicitly out of scope, such as third-party libraries you import unchanged, so nobody assumes they were checked.
2. Write the specification and threat model
Auditors check code against intent, so intent must be written down. Describe each actor (users, admins, keepers, oracles), what each is allowed to do, and the invariants that must always hold, for example "total shares never exceed total deposits" or "only the timelock can change fees". Then list what an attacker would want: draining a vault, minting without collateral, freezing withdrawals, or manipulating a price feed.
3. Reach meaningful test coverage
Line coverage is a floor, not a goal. Every external function needs tests for the happy path, every revert condition and the boundaries: zero amounts, maximum values, first and last user, empty pools.
Add fuzzing and invariant tests
Foundry's fuzzer and invariant testing, or Echidna and Medusa, throw thousands of random sequences at your contracts and check that your stated invariants survive. Invariant tests written from the specification are one of the strongest signals to an auditor that the team understands its own system.
Fork tests for integrations
If your protocol talks to an existing DEX, lending market or oracle, run tests against a mainnet fork so integrations are exercised against real deployed contracts rather than mocks.
4. Run static analysis and fix the noise
Run Slither and a linter such as Solhint, then fix or document every result. An audit report padded with findings an open-source tool could have produced wastes your budget and buries the serious items. Keep the compiler version pinned and the build warning-free.
5. Document the code itself
Use NatSpec comments on every public and external function, explain non-obvious maths, and keep a short architecture note showing how contracts call each other. A one-page diagram of the contract system saves auditors hours.
Common vulnerabilities to remove before review
Most serious findings fall into familiar families. Work through these yourself first:
- Access control gaps: unprotected initialisers, missing role checks on admin functions, or ownership that can be claimed by anyone.
- Reentrancy: external calls made before state is updated; follow checks-effects-interactions and use guards where needed, including read-only reentrancy through view functions.
- Oracle and price manipulation: reading spot prices from a single pool that a flash loan can move within one transaction.
- Rounding and precision: division before multiplication, share inflation on empty vaults, and rounding that consistently favours the user.
- Upgradeability mistakes: storage layout collisions between proxy versions and implementation contracts left uninitialised. Our guide to choosing a smart contract upgrade pattern covers the trade-offs.
- Unchecked external calls and token quirks: fee-on-transfer tokens, tokens that do not return a boolean, and rebasing balances.
Operational controls auditors will ask about
Code is only part of the attack surface. Expect questions on who holds admin keys, whether privileged actions sit behind a multisig and a timelock, how emergency pause works and who can trigger it, and how you will monitor contracts after launch. Have a written deployment script and a list of constructor parameters, because misconfigured deployments cause losses even when the audited code is sound.
Decide in advance how you will handle a critical finding found late: will you delay launch, reduce scope, or cap deposits at first? Making that call before the report arrives avoids pressure-driven shortcuts.
How to choose an auditor and plan the timeline
Book auditors early, because established firms often schedule engagements well in advance. Ask for sample public reports, the reviewers' experience with your chain and pattern (lending, AMM, bridge, NFT), how many engineers will review, and whether a fix review is included. For a deeper look at selecting a firm, read our buyer's guide to smart contract audit services in India.
Audit fees are driven by lines of code in scope, protocol complexity, novelty of the design, number of chains and how quickly you need the review. Preparation reduces all of the variable parts.
Where iJurug Soft fits in your audit preparation
We do not replace an independent auditor, and you should be wary of anyone who audits code they wrote. What our senior blockchain engineers do is build and harden contracts so they arrive at audit in good shape: specification and threat modelling, invariant and fuzz test suites, static-analysis clean-up, upgrade and admin-key design, and fixes after the report. It is part of our blockchain development services, delivered with fixed milestones from Discover through Launch.
Frequently asked questions
How long does audit preparation usually take?
It depends on the current state of tests and documentation. A well-tested codebase may need only a few days of clean-up; a codebase with thin tests and no specification needs considerably longer, which is why it pays to start before you book the auditor.
Is an automated scan enough instead of a manual audit?
No. Tools such as Slither catch known patterns, but the costliest exploits are usually logic and economic flaws specific to your design. Use tools to clear the basics so manual reviewers can focus on those.
Should we audit before or after testnet deployment?
Ideally, deploy to testnet first, run your own scenarios there, freeze the code and then audit. Any change after the audit should be minimal and re-reviewed.
Do we need a second audit after fixes?
Most firms include a fix review for the changes made in response to findings. Large redesigns after the report generally need a fresh review of the affected contracts.
Want a second pair of senior eyes on your contracts before the auditor starts? Tell us about your protocol, chain and target launch through our project form or write to info@ijurugsoft.com. We will review your repository, share a readiness checklist specific to your code and outline the work with clear milestones.