Strapi MCP Integration

Strapi is an open-source headless CMS that gives developers full control over their content API while providing editors with a friendly admin panel. The Strapi MCP integration in Weldable connects your AI agents to your Strapi instance, allowing them to create, read, update, and delete content entries through natural language. Manage your content pipeline without switching between the admin panel and other tools.

Use cases

Create content entries from external data

Your AI agent can read data from Google Sheets or Google Docs and create corresponding content entries in Strapi. This is useful for bulk content imports, product catalog updates, or migrating content from spreadsheets into your CMS.

Audit content for completeness

Ask your agent to scan your Strapi content types and flag entries that are missing required fields, have empty descriptions, or lack featured images. The agent compiles a report and posts it to Slack so your editorial team knows exactly what needs attention.

Publish and notify in one step

When content is ready to go live, your agent can change its status to published in Strapi and simultaneously notify your team via Slack or send a distribution email through Gmail. This combines two manual steps into a single command.

How it works

Connect your Strapi instance to Weldable by providing your API URL and authentication token on the integrations page. Your AI agent can then work with your content using requests like "create a new blog post in Strapi with title 'March Release Notes'" or "list all draft articles." Weldable maps your requests to the appropriate Strapi REST API endpoints.

Tips

Use content type names in your requests. Strapi organizes content by type (e.g., "articles," "products," "pages"). Include the type name so the agent knows which collection to target. This prevents the agent from guessing when your Strapi project has many content types.

Handle relations explicitly. Strapi content types often have relations to other types. When creating entries, specify related items by name or ID so the agent can link them correctly. One-to-many and many-to-many relations use arrays of IDs, and getting the format right avoids common API errors.

Test against a staging instance. If you run separate Strapi environments, connect the staging instance first to validate your workflows before applying them to production content. This is especially important for bulk operations that create or update many entries at once.