ArchitectureA supervised JVM-class runtime — OLTP on seven engines, OLAP on three. AI-native, MCP-native, observable as plain SQL.Read the architecture
Está viendo la edición Perú. Está viendo la edición Colombia. You're viewing the Pakistan edition. Cambiar a la edición global →Cambiar a la edición global →Switch to the global edition →
Product Release

MCP exposes data dictionary as schema resources and extends dict_validate coverage

The MCP server advertises the data dictionary as schema:// resources, loads schema content from the AI prompt catalog at runtime, and extends dict_validate to report, virtual table, and table object kinds.

This release extends the MCP server's schema-introspection surface in three directions: the platform data dictionary is now advertised as a set of schema://dict/* resources that MCP clients can browse and read; schema resource content is loaded from the AI prompt catalog at runtime so descriptions update without a server restart; and the dict_validate tool accepts report, virtual table, and table object kinds in addition to the existing set.

Dictionary schema resources

  • schema://dict/* resource tree. The MCP server lists dictionary schema resources in its platform capability block. Clients that request the resource list see dictionary schema entries alongside other platform resources, with each URI mapping to a schema description document.
  • Runtime-loaded content. Schema resource bodies are read from the AI prompt catalog at request time rather than compiled into the server binary. An administrator can update a schema description and the next client request returns the updated content without a deployment cycle.
  • Registered prompt codes. Platform-specific MCP prompt interactions are keyed by a registered identifier in the AI prompt catalog, making platform prompts enumerable and manageable through the same administration surface as other prompt entries.

dict_validate tool extensions

  • Report validation. dict_validate now accepts the report object kind, validating query bindings, column definitions, and output format declarations against platform schema rules.
  • Virtual table and table validation. Virtual table and table definitions are validated for column type consistency, key constraints, and relationship integrity, surfacing structural errors before the object reaches the application layer.
  • XSQL SELECT validation. SELECT-type XSQL statements embedded in dictionary objects are validated as part of the dict_validate pass, catching syntax errors and unsupported constructs before the object reaches runtime execution.

The dict_validate tool is callable from any MCP client, making it suitable for AI-assisted development workflows where a language model drafts an object definition and the tool validates it before the definition is committed. All extensions are additive; existing calls targeting previously supported object kinds continue to work without modification.

See the feature →

← All posts