Data Source Tools

Semantic Boundary

Data source tools own table-level capabilities:

• Retrieve a data source.
• Query a data source.
• Create or update a data source.
• List templates.
• Manage schema/properties.
• Create or update entries/pages through the Page/DataSource split.

data_source is not a synonym for database. A database is the container; a data source is the concrete table under that container. The current CLI/Core/MCP stack includes templates and property rename.

• data_source_retrieve
• Description: reads a Notion data source.
• Required parameters: data_source_id.
• data_source_query
• Description: queries a Notion data source.
• Required parameters: data_source_id.
• Optional parameters: payload.
• data_source_create
• Description: creates a data source.
• Required parameters: payload.
• Optional parameters: database_id, dry_run.
• Notes: when database_id is provided, calls DataSourcesService.create_for_database() and creates the data source under that database container; otherwise it uses raw payload create.
• data_source_update
• Description: updates a data source.
• Required parameters: data_source_id, payload.
• Optional parameters: dry_run.
• data_source_templates
• Description: lists page templates available to a data source.
• Required parameters: data_source_id.
• data_source_property_rename
• Description: renames a data source property by property ID or current property name.
• Required parameters: data_source_id, property_id_or_name, new_name.
• Optional parameters: dry_run.

MCP data source tools call DataSourcesService. They do not call CLI and do not import the Notion SDK directly. Table-level query/schema/templates/property operations must not be placed under database container tools.