- SCHEMA
- DATA
- CMS
- HUMAN
Short answer
How can I use AI to build CMS content models?
Your agent inspects your current Cosmic schema and uses that logic to create new object types. It then generates and pushes JSON payloads to populate those objects. You get a structured CMS ready for use.
The results
Where your data lands
This is what happens once your agent starts working with Cosmic.
SCHEMA
Mapped object types
The AI identifies your existing data structures. You get a clear blueprint of how new objects should look.
DATA
Populated JSON payloads
Your agent converts raw text into valid JSON. These payloads are ready to be injected into your CMS.
CMS
New object types
The AI creates the actual containers in Cosmic. Your new content types appear in your dashboard immediately.
HUMAN
Final validation
You review the created objects and the data accuracy. The AI does the heavy lifting, but you keep the final say.
The workflow
What your AI does when the request arrives.
The agent moves from inspecting your current setup to executing the new data structure.
Audit existing types
The agent looks at what you already have in Cosmic to ensure the new models match your current naming and structure conventions.
list_schema_typesAnalyze specific fields
To get the details right, the agent pulls the specific requirements for your target schema.
get_schema_detailsGenerate payloads
Your agent takes your raw input and formats it into the exact JSON structure required by the CMS.
get_schema_detailsCreate the objects
The agent pushes the finalized data and the new object definitions directly into your Cosmic environment.
create_cms_object
Try it
Copy these to start.
Use these prompts to kick off the modeling process.
Starting points
These are opening lines. Swap out the bracketed text with your specific CMS names or data sources.
Cosmic Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_8AgjmeIFMLOZAgK2CfQoBzmnSYL11jJUw3E3Nh4a/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — Cosmic capabilities are ready to use.
{
"mcpServers": {
"cosmic-mcp": {
"url": "https://edge.vinkius.com/vk_preview_8AgjmeIFMLOZAgK2CfQoBzmnSYL11jJUw3E3Nh4a/mcp"
}
}
}Look at my current Cosmic schema and suggest a new object type for [blog posts].
Use the details from my [product] schema to create a JSON payload for a new entry.
Check my existing [user profile] types and then create a new object for [customer reviews].
Generate a JSON payload for a [news article] based on this text and create the object in Cosmic.
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect Cosmic once, then ask.
Just click the link to connect your account. Your credentials stay encrypted, and you can immediately start prompting your agent.
Connect Cosmic to your AIFAQ
How this task behaves
- 01
Can the AI delete my existing content?
No. The tools provided only allow for listing, getting details, and creating new objects. It cannot delete or modify existing data.
- 02
Does the AI decide the field types for me?
The AI suggests field types based on your existing schema, but you should always verify the structure before the agent executes the creation step.
- 03
Can I use this to update old entries?
No. This specific workflow is designed for creating new object types and new objects. It does not have the capability to update existing records.
- 04
Do I need to write the JSON myself?
No. Once the agent understands the schema, it generates the JSON payloads for you.
- 05
Will the AI change my existing schema settings?
No. The agent can read your schema to learn how to behave, but it cannot change your existing configuration.
More questions about Cosmic? See everything the Cosmic Connector can do
Connect Cosmic to Claude, Cursor, ChatGPT & more
