Security

Secure WordPress MCP starts with scoped access.

MCP gives AI agents real capability. MCPWP should be introduced with clear scopes, read-first inspection, revocable credentials, and explicit handoff points before risky changes.

Operating model

The safe path is inspect, prove, then write.

Read-only first

Connect the client and verify site info, active capabilities, and available tools before any content or layout changes.

Scope by workflow

Use the narrowest key that matches the job: content, Elementor, SEO, media, menus, or admin operations.

Revoke when done

Treat API keys like deployment credentials. Rotate or remove them after short-term external access.

Track blockers

When a workflow exposes a plugin bug or missing feature, create an issue so launch and product work stay connected.

Trust copy

Security claims must be concrete.

Do not claim magic AI safety. State the actual controls: endpoint authentication, scoped keys, read-only setup, logs where available, and operator review before destructive actions.

Authenticated endpointExplicit MCP client configurationScoped API key setupHuman-reviewed write operationsStaging recommended for risky layout or code changes
FAQ

Questions buyers ask before connecting AI to WordPress.

Can an AI agent damage a site?

Any write-capable integration can cause harm if over-permissioned. MCPWP messaging should always recommend read-first setup and scoped keys.

Is staging required?

For content and metadata, production workflows can be scoped. For risky layout, code, commerce, or broad automation work, staging is the safer default.

What should agencies standardize?

A key policy, first-connection checklist, rollback path, issue reporting path, and approval rules for destructive operations.