WordPress security audit: checklist before hiring maintenance
Published on 2026-11-14 by frfrand
What to check in a WordPress audit: versions, access, backups, permissions, vulnerabilities, and provider obligations.

# WordPress security audit: checklist before hiring maintenance
Hiring WordPress maintenance should not mean paying for someone who only clicks "Update plugins" once a month. A company needs to know what risks it has, how the website will be restored, and who will respond to an incident. This list allows evaluating a provider based on evidence, not promises.
Verifiable technical inventory
Request an updated list of the core, PHP version, database engine, active theme, installed plugins, and origin of each component. Mark abandoned, unsupported, duplicated, or unnecessary elements. Document external services used for forms, payments, email, and analytics. Expected test: inventory with versions and review date, not a generic screenshot of the panel.
Identities and permissions
Check administrator users, former employee accounts, and multi-factor authentication. Review access to hosting, SFTP/SSH, database, DNS panel, and email accounts. A provider does not need to share a single administrator password across its entire team. Expected test: list of authorized roles and onboarding, offboarding, and emergency procedures. Avoid delivering secrets in reports.
Updates and vulnerabilities
Review official advisories, relevant CVEs, patched versions, and whether components are actually exposed. Risk is not decided solely by a CVSS score: required privileges, affected function, known exploitation, and available mitigations matter. Agree when to apply critical patches and how they are tested. Expected test: change log with previous version, new version, and test result.
Recoverable backups
Ask where file and database backups are stored, how many versions are kept, if they are encrypted, and if a compromised account could delete them all. Require at least one periodically tested restoration; a backup that has never been restored is a hope, not a guarantee. Define RPO (how much data can be lost) and RTO (how much recovery time is sought) in terms aligned with the business.
Configuration and infrastructure
Evaluate HTTPS, certificates, secret protection, file permissions, service exposure, logs, and system updates. Check that database access is restricted and that the server separates applications when appropriate. A WAF helps, but does not justify keeping a vulnerable plugin.
Monitoring and response
The provider must indicate what is monitored: HTTP availability, 5xx errors, resources, suspicious changes, and certificate expiration dates. Define reporting channel, responsible parties, service hours, severity, and escalation.
| Area | Minimum evidence |
|---|---|
| Software | Inventory and patch plan |
| Accesses | Audited users and roles |
| Backups | External copies and tested restoration |
| Incidents | Containment plan and contact |
| Changes | History and rollback |
Contractual red flags
Be wary of claims of absolute security, backups without testing, shared access without traceability, or lack of responsibility for tasks. Request an explicit separation between preventive maintenance, functional improvements, and incident recovery.
Frequently asked questions
### Does an audit replace maintenance? No, it is an initial diagnosis; risk changes over time. ### Is it necessary to review every plugin? Yes, especially those exposed and handling sensitive data. ### Is hosting the developer's responsibility? It depends on the contract; it must be in writing.
To continue
Complement with infrastructure security and what is a CVE. If you hire ArduMaker services, define scope, backup review, and incident response in writing. Technical source: WordPress Security Hardening.