Everything the laboratory publishes openly, hosted on Google Drive. The document database is the raw material behind most of this site — it is offered as-is, unedited, including the parts that disagree with each other.
The full document corpus, organised into numbered subject folders — catalogues, materials, space and thermal, displays, robotics, bioids, MycoLink, mRNA therapeutics, Rodin coil, DARPA and patent material, and cross-project summaries. This is the primary source behind the technology pages on this site.
Two honest warnings for anyone reading it. The corpus spans about a year and the later documents supersede the earlier ones without the earlier ones being withdrawn, so figures conflict in places. And it includes a worldbuilding folder alongside the technical work — that material is fiction and is not labelled as such inside the folder structure.
A Windows desktop application built by the laboratory. The folder holds the signed installer alongside the archived builds behind it, so an earlier version can be retrieved if a newer one misbehaves.
The reference skill bundle that encodes the laboratory’s canonical vocabulary, figures and evidence discipline for use with an AI assistant — so a model working on Hardin Labs material uses the current names and numbers rather than inventing them, and carries the simulated-versus-measured distinction through.