Fix: Support schema-qualified table names in column retrieval

[tripathishivani]

The Problem
The current implementation of the run method in the SQL Server MCP node has two primary issues when retrieving table columns:

1. Schema Ignorance: If a user provides a schema-qualified name (e.g., sales.Customers), the current query fails to return results because it searches for the literal string sales.Customers within the TABLE_NAME column of INFORMATION_SCHEMA.COLUMNS.
2. Namespace Collision: If the same table name exists in multiple schemas (e.g., dbo.Users and archive.Users), the current logic could return a combined list of columns from both tables, leading to data corruption in the MCP output.

The Solution
This PR updates the table name parsing logic to be schema-aware:

1. Parsing: It now splits the input tableName string into schema and name components.
2. Defaulting: It defaults to the dbo schema if no specific schema is provided, ensuring backward compatibility.
3. Query Accuracy: The SQL query has been updated to filter by both TABLE_NAME and TABLE_SCHEMA.

Changes

1. Updated run function in MssqlMcp/Node/src/tools/DescribeTableTool.ts
2. Modified SQL query to use @tableSchema parameter.

Added logic to handle both simple (Users) and qualified (dbo.Users) table names.