I use it now in almost every application I do :-)
My way to work is to combine the shell script with it,
the result is amazingly fast and clean :-)
but I "recently" solved what I imagined as the weakness of the Sqlite command line:
it is the CSV import !
in real work life, we need to treat CSV files like:
my field 1,"my opinion, my view on it", my field 3
or simply a field on multiple lines like
"my
multi-lines
field",...
the .import does not manage this kind of CSV file....
I wrote a state machine (a big switch/case) to parse the CSV file.
I progress in my state machine by getting the next character until the end of the file.
The funny thing is that I always try to rely on ".import" when I know I am using simple files,
but I had to change it for one file, the day after I wrote my own parser :-)
funny isn't it?
long life to SQLite !