The scripts, docs and workspace dependencies here still described a personal data server, a hook, a swarm and a setup binary, none of which are in this repository. What is left is the read side: an index, a query service and a canvas, pointed at a server from elsewhere or at a recording. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Change-Id: I5fe14cf1371c16a4bffb9ca4bb1671691f0e05ec