Here, we describe the synthetic molecular evolution of a family of host-compatible antimicrobial peptides (AMP) designed for the treatment of wounds infected with drug-resistant, biofilm forming bacteria. The peptides tested are variants of an evolved AMP called D-CONGA, which has excellent antimicrobial activities in vitro and in vivo. In this generation of rational D-CONGA variants, we tested multiple sequence-function hypotheses that had not been tested in previous generations. Many of the peptide variants tested have lower antibacterial activity against Gram-positive and Gram-negative pathogens, especially those that have altered hydrophobicity, secondary structure potential, or spatial distribution of charged and hydrophobic residues. Thus, D-CONGA is generally well tuned for good antimicrobial activity. However, we identified one variant, D-CONGA-Q7, with a polar glutamine inserted into the middle of the sequence, that has higher activity against both planktonic and biofilm-forming bacteria and lower cytotoxicity against human fibroblasts. Against clinical isolates of K. pneumoniae, innate resistance to D-CONGA was surprisingly common despite a lack of inducible resistance in P. aeruginosa reported in our previous work. Yet, these same isolates are susceptible to D-CONGA-Q7. Additional mechanistic work will be required to understand why D-CONGA-Q7 is less prone to innate resistance in Gram-negative bacteria. In the spirit of synthetic molecular evolution, the discovery of D-CONGA-Q7 achieved a critical gain-of-function, and has provided a significantly better template sequence for the next generation of synthetic evolution.