Skip to main content
Metabind lets your agent answer with your own UI components, in Claude, ChatGPT, every MCP host, and natively in your apps. Content is one of the things those answers can draw on. A content project holds content types, content entries, and assets that your team edits in Metabind Studio, your apps query, and your agent reads through a Data Tool and shows through an Interactive Tool. This tab applies to content projects only. To build the tools your agent calls, start with the Guides. This tab covers how to model content, how to manage it, how to query it from your applications, and how an MCP App uses it.

​
What “content” means in Metabind

These map to the headless CMS shape many teams already know. The difference is what they connect to: Metabind’s content can be rendered by BindJS components in your apps, served through REST and GraphQL APIs, and read by your agent through tools.

​
Where to manage content

In Metabind Studio:
  • Content tab in the project header — list, search, create, edit, publish content entries.
  • Types tab — define and version the schemas content entries follow.
  • Assets tab — upload, manage, and reuse assets across entries.
Editors can work without touching code. Engineers can use the REST or GraphQL APIs for bulk operations.

​
Where to consume content

Three pathways: For an agent, a Data Tool queries Metabind’s GraphQL API to fetch content entries, and an Interactive Tool renders them in the conversation: inside the host’s frame in Claude and ChatGPT, and natively in your own apps. This is how the same content backs both your app and your agent.

​
Content vs. Types/Tools

Metabind has two related-but-distinct registration units:
  • Tools are Interactive Tools or Data Tools the agent can call. See Tools and Types.
  • Content types are content schemas — content entries your agent or your app can read and write.
Each project is either a content project or an MCP project. Content projects show the Content and Types tabs, where Types holds content types; MCP projects show the MCP tab instead, which lists their MCP Tools. They interact through Data Tools: a Data Tool can read content through the Content API, or write it through the Admin API, giving the agent controlled access to your content.

​
Content delivery APIs

All three speak the same content model. Pick whichever fits your client.

​
AI-driven content workflows

Content can be created and updated by AI agents — useful for:
  • Drafting articles or product descriptions with AI assistance, edited by a human, published when ready.
  • Bulk migration of content from another CMS using an AI agent that maps fields.
  • Operations like archiving stale entries, tagging by topic, or normalizing metadata.
The content-management MCP server exposes context, discovery, and creation tools. See AI content creation.

​
Where to go next

Managing content

Editor workflow in Metabind Studio.

Querying content

REST, GraphQL, and SDKs for client apps.

AI content creation

Using AI to draft, modify, and migrate content.

Mobile SDKs

Native iOS and Android content fetching with BindJS rendering.