Overview

The Agiler MCP server connects an AI assistant to your Agiler account through the Model Context Protocol. You describe a task in conversation; the assistant calls tools to read data or make changes and explains the results.

What you can do

With the appropriate permissions, an assistant can:

• Inspect projects, domains, environment variables, and request rules.
• Read application and request logs, and review usage.
• Read and edit project files.
• Run SQL and WP-CLI commands.
• Create backups, manage backup schedules, and restore a site.
• Create, update, or delete projects.

The tools use the same REST API as other integrations. Account billing and workspace membership changes remain dashboard or API tasks. See the MCP reference for the complete catalog.

Choose your assistant

All direct connections use https://mcp.agiler.io/v1 and an Agiler API key sent as a bearer token. Agiler hosts the server; you do not need to run a local MCP process. Client features and organization policies determine where you can add the connection.

Decide what the assistant can access

An API key controls both the operations available to the assistant and the workspaces or projects it can reach. Your Agiler account’s permissions still apply. Start with a key restricted to one project and the scopes needed for your task.

MCP actions affect the hosted site. Editing a file through MCP is a live change, even when the conversation takes place inside a local code editor. Prompts and client approval settings guide the assistant; API-key permissions enforce access.

Get started

1. Create an API key and choose permissions.
2. Follow your client’s setup guide above.
3. Complete your first task to verify the account and project.
4. Try an example workflow and add project instructions for recurring work.