Abstract
Retrieval-augmented generation solved the wrong half of the memory problem. It made relevant text available to a model; it did not make the retrieved text accountable, and in the standard arrangement the retrieved passage is concatenated into a prompt where it is indistinguishable from instruction. This paper describes Mnemos, the memory extension of our convergence substrate, which takes the opposite position: a recalled item is a proposal, carrying its provenance, subject to the same promotion gate as any other proposal, and incapable of becoming a fact without passing it. Two consequences follow. The first is a security property that falls out of the architecture rather than being added to it: retrieved text cannot acquire authority, so indirect prompt injection reaches the context as an attributed proposal that policy may refuse. The second is a retrieval result. Mnemos ranks vector similarity and BM25 separately and merges the rankings by reciprocal rank fusion rather than merging the scores, and we show why this is not a tuning preference: the two scores live on incomparable scales, and rank fusion is invariant under any monotone rescaling of either, which score fusion is not. We close on the tension we have not resolved — an append-only substrate and a memory that must forget are not obviously compatible, and we state the problem rather than claim it solved.