Adding (and) logic for a list of columns and generating feature

I have requirement to create a feature by checking the fields of two different dataframes.
df1 is a dataframe created from mysqldb and it will be static, the values of column(required) from df1 can be moved from True or False.

below are more details.

df1= 
flag	beta	required
alpha	A	     TRUE
alpha	B	     TRUE
alpha	C	     FALSE
alpha	D	     TRUE

df2= 
name	A	B	C	D	E	F
roy	    1	1	0	1	0	0
john	0	1	1	1	0	0
sam	    1	1	1	1	1	1

I trying to create feature alpha_flag by using the column name "beta" from df1 if the required column is TRUE.

So the code for alpha_flag will be like below

df2['alpha_flag'] = np.where(df2['A']==1 and df2['B']==1 and df2['D']==1 , 1 , 0)

The difficulty i am facing is to choose only the columns that are mentioned as required true from df1 and dynamically update my 'alpha_flag' creation condition.

IIUC, you can use boolean indexing to select the beta rows where required is TRUE, then check if these columns in df2 is all equal with 1

df2['alpha_flag'] = df2[df1.loc[df1['required'].eq('TRUE'), 'beta'].tolist()].eq(1).all(axis=1).astype(int)

Note that if all required value is FALSE, this will cause selecting empty column from df2 and lead to empty dataframe manipulation which might cause unexpected behavior, so a more robust version would be

cols = df1.loc[df1['required'], 'beta'].tolist()
df2['alpha_flag'] = df2[cols].eq(1).all(axis=1).astype(int) if len(cols) else 0