It takes advantage of the PyRanges package (Stovner and Sætrom, 2019) and uses several of its functions to efficiently parse the sequence of thousands of transcripts and return its extended ORFs.
You can install extend_orfs using PyPI or download it directly from GitHub.