Working with AI

Last updated: July 3, 2026

This step covers how to combine an LLM agent with WEEGLOO MCP to build a content-based service faster.

WEEGLOO exposes most of its operations for external use through the CMA (Content Management API): defining a Content Type, creating and publishing Content, hosting, and more. It then makes these tools directly available to an LLM agent through MCP. With this setup, much of the work people used to do by hand in the content studio can be carried out on your behalf through natural-language requests alone.

The two examples below walk through, step by step, the concrete flow of putting an LLM agent and WEEGLOO MCP to work.

Migrating a static page with AI

This example covers the process of moving a static HTML page you have already built into a content-based service managed by WEEGLOO. You can see how an LLM agent handles the entire flow, from analyzing the page, defining the Content Type, loading the Content, through to integrating with the CDA, using natural-language requests alone. If you want to turn a static page already in production into a content-based service, start with this example.

Building and deploying a site with AI

Going one step beyond the example above, this one covers building a new site and making it publicly available. It walks through the full process: finishing a page's design and implementation with an AI-based design tool, working through content management and CDA integration, and finally publishing it externally with WEEGLOO Web Hosting. If you want to build a new service from scratch, or complete the whole path from design to deployment using only an LLM agent and WEEGLOO, see this example.