Science One Framework: Verifiable Autonomous Research via Chain-of-Evidence

Interesting question is if the CoE Audit evaluates evidence chains recursively (each cited paper's own evidence chain), or stops at citation existence + method-code alignment for the current paper.

For example, citing a paper whose methodology gets refuted by follow-up literature.