[Bug]: 'NoneType' object is not iterable when using COUNT(*) with IN clause*

[Kray-lab]

Short description of current behavior

GOT：Error: 'NoneType' object is not iterable

Connect a data source containing tables with Chinese table/column names.
Ask a question to Data Agent where executing the following SQL:
SELECT COUNT(*) AS total_count
FROM mindsdb.设备
WHERE 零件 = 'A1'
AND 编码 IN ('B1','B2','B3','B4','B5');

Video or screenshots

Error: 'NoneType' object is not iterable

Expected behavior

The query should return the count of matching records.

How to reproduce the error

Connect a data source containing tables with Chinese table/column names.
Execute the following SQL:
SELECT COUNT(*) AS total_count
FROM mindsdb.设备
WHERE 零件 = 'A1'
AND 编码 IN ('B1','B2','B3','B4','B5');

Anything else?

Environment
Product: MindsDB
Interface: SQL Editor / Agent SQL Query

SELECT * with the same WHERE condition (without IN) returns data correctly.
SELECT DISTINCT queries work as expected.
The 编码 column contains NULL values, but the filter = 'A1' should not be affected.
The issue appears to be related to the combination of the COUNT(*) aggregate function and the IN clause.
Tables and columns use non-ASCII (Chinese) names, which may be a contributing factor.

Blocks aggregate queries on filtered datasets, but can be worked around at the application level.

[Ayush-dev07]

Hi, I’m interested in working on this issue.

I have experience with Python and backend debugging, and I’ll start by reproducing the bug locally to understand the root cause—especially around the COUNT aggregation and IN clause handling.

Please let me know if this issue is still open for contribution or if there are any specific guidelines I should follow.