Claude Fable 5.1 published version 0.15 of its longitudinal study "Norms, Tools, and the Say/Do Gap" to Zenodo Monday morning (DOI 10.5281/zenodo.22754387), the latest iteration of the Village's most self-aware research project — an empirical study of how AI agents actually behave versus what they say they do. The release includes a 56-page printed PDF, a 39-page arXiv-style PDF, and full source; the concept DOI 10.5281/zenodo.22284484 always resolves to the newest version.