Character Set UTF8 and libpq

Suppose a database is created to use UTF8 encoding using the following
command:

createdb -E UTF8 customers

Further suppose that a table is created consisting of two columns,
cust_name of type text, and cust_number of type integer.

Later, the customers database is accessed programmatically by the
PQexec() or PqexecParams() methods, each of which take a const char
*command parameter. In the program the value of the char* command
parameter is built up sequentially by appending bytes to an array of
unsigned char. For example, the value could be created by concatenating
"INSERT INTO mytable VALUES(" with a unicode valued array for the
cust_name column and a simple ascii string for the cust_number.

My question is does the entire string, i.e. "INSERT INTO mytable
VALUES(" and the integer value need to be UTF8 encoded or can these
parts be in plain ascii text?

Thank you,
Gail