A lab that keeps the notebook

A small wet-lab and methods group stopped scattering findings across chats, shared drives, and one person’s laptop. The notebook lived in a vault they owned. Agents could draft a protocol revision or a literature digest; a named investigator had to accept it before it became the record.

Reproducibility improved because refusals were visible: which draft was rejected, who signed the version that ran, what the instrument log actually said. Visitors could review without harvesting the lab’s unpublished life.

Research memory should compound. Soft outcome: the science still moves; the notebook does not vanish when a postdoc leaves or a chat thread scrolls away.