ADR-013: fold the forked substrate into alkstore-sqlite — no fourth crate

Operator review of ADR-012's fork design re-litigated §1's crate
identity. alkstore-substrate misdescribed what the code is (unpublished,
path-dep-only, one consumer, SQLite-only — not a family-wide substrate);
the mechanical-diff hope was gone at fork time regardless (port deltas,
renames, re-derived half); and the alksocks F-1 lesson applies — a
vendored region under a second, weaker instruction set is a defect seam.
The fork folds into alkstore-sqlite as a bounded module subtree
(src/substrate/); ADR-012 §3–§6 retained verbatim, §2 retained with its
enforcement re-sited from the crate graph to diff fence + review +
contract-suite equivalence pins. OQ-11 item (1) dissolved.

docs/architecture/decisions/003-sqlite-driver.md

@@ -41,7 +41,9 @@ The SQLite engine uses **posture 1**: published `honker-core = 0.5`
 as a library dependency over the crate's own rusqlite connections
 (`bundled-sqlite` for hermetic builds). *(Ownership superseded
 2026-10-05 by [ADR-011](011-sqlite-substrate-fork.md): the substrate
-is the forked `alkstore-substrate` lineage crate, not the published
+is the forked honker-core lineage, carried in-tree as the engine
+crate's substrate module subtree per
+[ADR-013](013-fold-substrate-into-sqlite.md), not the published
 dependency — the driver, seam, and watcher architecture below are
 unchanged.)*