sanity-mcp-server
Sanity MCP Server is designed to enhance content management by integrating AI capabilities with Sanity projects. It uses the Model Context Protocol to enable AI tools to interact with and manipulate content through natural language commands, supporting various applications and enabling efficient content operations.
What is the Model Context Protocol?
The Model Context Protocol (MCP) is a standard that allows AI models to interact with content management systems like Sanity, enabling them to understand and manipulate content structures through natural language.
How do I configure the Sanity MCP server?
You need to add the server configuration to your application's MCP settings, specifying the command, arguments, and environment variables such as project ID, dataset, and API token.
What permissions are required for the API token?
The API token should have appropriate permissions based on your usage. For basic read operations, a 'viewer' role is sufficient, while 'editor' or 'developer' roles are recommended for content management.
Can I use the MCP server with production datasets?
Yes, but be cautious as the AI can perform destructive actions like creating, updating, or deleting content. It's advisable to use a development or staging dataset for testing AI operations that require write access.
What are the supported user roles?
The server supports two user roles: 'developer' with access to all tools, and 'editor' with content-focused tools without project administration capabilities.