The interplay between DNA and proteins extends beyond the transcription of protein coding genes. Proteins can return to bind DNA sequences, often with high specificity and precision. One example of DNA-binding proteins are RNA-guided DNA nucleases of clustered regularly interspaced short palindromic repeats (CRISPR) systems. Originally evolved as a prokaryotic immune system against invading bacteriophages and other mobile genetic elements, CRISPR systems and their associated (Cas) nucleases have revolutionized the fields of biology and medicine. Despite their widespread use as genome editing tools, there are two major limitations to their use in the clinic. First, commonly used viral delivery vectors have constrained packaging limits that are exceeded by the most popular nucleases, such as SpCas9. Second, genomic targeting of a Cas nuclease of interest is determined by its unique recognition profile of a specific DNA motif. Therefore, long and complex PAMs greatly restrict the availability of genomic sites targetable for gene editing. Here, we address these limitations with directed evolution and protein engineering methods to broaden the DNA recognition profile of miniature RNA-guided DNA nucleases. $\textit{Full-text embargoed until: 2027-06-29}$