Improve Microsoft SQL node (e.g. schema, table, column autodetection and mapping)

Using the Microsoft SQL and Postgres node side by side, the Postgres node has in my opinion a better usability. The Microsoft SQL node lacks some useful features.

I’m currently using n8n 1.55.3 embed.

I think the Microsoft SQL node could be improved with features from the Postgres node:

  • Detection and dropdown selection of schema and table
  • “Mapping Column Mode” is missing completely:
    • “Map Automatically” for Postgres is very powerful, where Microsoft SQL node requires configuring all field names, which fields must also be exactly named in workflow (usually requires “Edit Fields” node)
    • “Map Each Column Automatically” with automatic column detection in Postgres also very helpful and supports expressions
  • Microsoft SQL node has no “Select” or “Insert or Update” query built in
  • Postgres supports “Execute Query” with query parameters to prevent SQL injection attacks. Is there a workaround for the Microsoft node or is it actually vulnerable to SQL injections?

n8n would benefit from all these improvements, but if I would have to choose, I would definitely love to see mapping columns with expressions and query parameters for “Execute Query” first.

Thanks!

Hi @pls ,
Thanks for the great input! :raised_hands:

To answer your question about the SQL queries, there isn’t currently a workaround for the MIcrosoft node, but we do have it our roadmap to overhaul our SQL nodes to abide by prepared statements to avoid issues of this nature.

Thanks for the feedback @mariana-na!

Glad to hear it’s on the roadmap. May I ask for a rough estimate?

Not before the new year I’m afraid :slight_smile: