Standardize JSON Output Schemas for Common Data Entities

[rgeissen]

Is your feature request related to a problem? Please describe.
I'm always frustrated when I have to write complex, defensive parsing logic on the client side because different tools return the same kind of information using inconsistent JSON keys. The provided agent log highlighted a bug where the _execute_column_iteration function failed to correctly parse a column's data type because it was looking for the wrong key in the JSON returned by the metadata tool. This makes client-side integration brittle and prone to breaking whenever a tool's output schema is slightly modified.

Describe the solution you'd like
I would like the MCP server to enforce a standardized and consistent JSON schema for common data entities across all relevant tools. When multiple tools can return information about the same entity (e.g., a database column), they should use the exact same structure and key names.

For example, any tool returning column metadata (qlty_columnSummary, base_columnDescription, etc.) should always use a consistent, case-sensitive key for the data type, such as dataType.

Example of Standardized Output:

// From tool A
{
  "ColumnName": "ZIPCODE",
  "dataType": "INTEGER", 
  ...
}

// From tool B
{
  "ColumnName": "ZIPCODE",
  "dataType": "INTEGER",
  ...
}

[remi-td]

Agreed. Perhaps the best first step would be to extract and maintain somewhere a catalog of the terms used in tools i/o, so we can start standardising. Happy if you want to get started, @rgeissen