Is it safe to feed the same adapter list for both forward and reverse strands (for paired end reads)? In other words, will the tool automatically reverse the adapters for the reverse strand?

cutadapt --quiet -q 10,10 -aA /Volumes/My\ Passport/Documents/adapters.fa -o out_1.fastq -p out_2.fastq in_1.fastq in_2.fastq

I purposefully left out -O command (min overlap length, of 25) because the overrepresented Kmers in my FastQC report showed 6 bp sequences commonly found in TruSeq adapters. The default min overlap is 3 bp.

Also, in an example pipeline, I found these (highlighted) commands that I could not understand why it was necessary.

ls *R1_001.fastq | parallel -j $PARALLEL_TASKS "cutadapt -q 10,10 -a GATCGGAAGAGCACACGTCTGAACTCCAGTCAC -O 25 {} 2> {.}.cutadapt | sed 's/^$/C/g' > {.}_adapt.fastq​
                                                                                                             ^^^                      ^^^^^^^^^^