index

Katz 0.587

The ideas graph indexer: parse `.rq` with the native engine, extract ideas / edges, persist to rusqlite SQLite.
Same-file `member` edges target the member leaf ( references_to_subidea ).
Wildcard path + idea fan-out: wildcard_reference_edges
( wildcard_references ).
Ideas Summary graph can hide those edges via webview.rq /
wildcard_refs_toggle.
Soft mtime sync skips unchanged files. `extract_version` in the ideas index meta is the extract-rule generation. When that number does not match the engine, sync rebuilds so stale edges do not remain ( inline_code ).
Owned by ownership; consumed per consumption.

tests

schema.rs:opens_at_schema_version_4, schema.rs:schema_v4_columns_round_trip_git_count_and_mtime, wildcard_reference_edges.rs:sync_persists_wildcard_reference_edges, extract_parity.rs:qualified_ideaset_member_reference_targets_leaf_idea_not_ideaset, broken_references.rs:sync_drops_stale_inline_code_file_refs_when_extract_version_changes

Interactive page: index