This is a harvest manifest, not a deep-dive: every vespa/ top-level directory not yet
referenced by any governed note’s sources: was enumerated mechanically (directory listing,
cross-checked by analyze_garden.py) and classified into a MOC family by name-level
evidence plus a directory peek (CMakeLists.txt/pom.xml/README presence), not by
reading module internals. Enumeration is machine-checked; the family classification is project-inferred (hence corpus-level confidence: low). Each family row names the target
MOC and the candidate next-depth note (zettel/bridge/guide) for the follow-up harvest waves.
Family Map
Config & Control Plane (target MOC: config-admin, GDN-NOTE-0004)
These are Maven/CMake aggregation or test-support artifacts. Decision recorded: they
count as covered by classification once this note lands; no zettels will be minted for
them individually. Re-visiting requires a contradiction-register entry.
Next Waves
Wave A (config plane): the config/zookeeper/slobrok family -> 2 zettels + 1 bridge.
Wave B (container assembly): container/container-dev/container-search-and-docproc
aggregator-vs-code resolution -> resolves an open question in container-jdisc.
Wave C (storage plane): persistence/storageserver/vdslib/streamingvisitors.
Wave D: split a dedicated mocs/linguistics.md if the family grows past ~6 notes.