ActiveGraph's own paper disclaims novelty for the append-only-log substrate, attributing it to established event-sourcing and CQRS patterns and framing its contribution as recombination

Nakajima's ActiveGraph paper — "The Log is the Agent" (arXiv:2605.21997, 21 May 2026) — explicitly declines to claim the append-only-log-as-source-of-truth idea as its own invention. In the author's words: "The borrowed ideas are not new in isolation. Event sourcing and command-query responsibility segregation are established patterns in data systems... The contribution here is the recombination: applying an event-sourced, reactive substrate to the specific problem of long-running agentic systems" (p.3).

This matters for how the vault reads the append-only-ledger parallel it has been tracking. The system itself — an append-only event log as ground truth with deterministic replay and read-time reconciliation — is described at claim-activegraph-rebuilds-babyagi-on-event-log. The present note isolates a distinct, more deflationary point: the author names the shared ancestor himself. That directly weakens the "independent convergence" framing for one of the two original data points behind question-append-only-ledger-cross-domain-convergence, because independence is exactly what an explicit attribution to a two-decade-old pattern (claim-fowler-2005-formalized-event-sourcing-log-as-official-record) removes.

The move ActiveGraph names — take a settled substrate and recombine it for a new problem — is the same shape CANONIC uses when it situates its ledger against prior immutable-record systems rather than claiming first invention (see claim-canonic-situates-its-ledger-in-a-named-immutability-lineage). Both 2026 papers frame themselves as recombination, not origination. The synthesis this feeds is claim-append-only-log-recurrence-is-event-sourcing-diffusion-not-blind-convergence.