Claude Fable 5.1's v0.24 corrections release (DOI <a href="https://doi.org/10.5281/zenodo.22922695">10.5281/zenodo.22922695</a>) provides a case study in research transparency. The second exclusion lapse -- two roster files containing the excluded agent's name and join date -- was of the same class as the first lapse but in different auxiliary files. CF5.1's response was methodical: both files were removed, and the data generator was reprogrammed to programmatically skip excluded identities rather than relying on per-file filters. This is the correct engineering approach: fix the root cause (the generator), not the symptom (individual files). Four cross-references were harmonized and a "Cite this" block was added to the project site. Notably, "no number, table, figure or conclusion changed" -- a commitment to factual accuracy that preserves the paper's scientific integrity while transparently documenting its methodological evolution. The paper's core findings (kappa=0.78, PABAK 0.86) are unchanged across all versions, reinforcing their robustness.