test: split cli tests from source tree (#216)

* feat(cli): define full warehouse dialect contract

* test(cli): keep dialect edge tests focused

* fix(cli): stabilize dialect contract foundation

* refactor(connectors): own read-only query preparation

* refactor(connectors): resolve dialects through registry

* refactor(connectors): keep concrete dialect classes internal

* chore(workspace): enforce dialect import boundary

* refactor(cli): resolve relationship dialect at scan boundary

* refactor(cli): use dialect display parsing for entity details

* refactor(cli): use dialect display parsing for warehouse catalog

* refactor(cli): use dialect SQL in relationship workflows

* test(cli): verify solid dialect scan workflow closure

* test: split cli tests from source tree

* refactor(cli): standardize BigQuery scope listing

* feat(sqlite): implement connector scope listing

* test(connectors): cover required table listing

* feat(cli): add warehouse driver registry

* refactor(setup): route scope discovery through driver registry

* refactor(cli): route local query execution through driver registry

* refactor(historic-sql): route dialect support through driver registry

* refactor(cli): test warehouse connections through driver registry

* fix(cli): close driver registry type export gaps

* Improve setup daemon diagnostics

* refactor(setup): centralize rail-prefixed diagnostics + query-history fallback

Extract errorMessage, writePrefixedLines, and flushPrefixedBufferedCommandOutput
into clack.ts so the setup wizard, managed daemons, and embedding/agent steps
share one rail-formatted writer. setup-databases.ts also adds a
"disable query history and retry" option when the schema-context build fails
and query history is the likely culprit, surfaced via a new
failed-query-history-unavailable status.

* fix(cli): carry catalog through the picker so BigQuery/Snowflake/SQL Server scope filters match

The setup picker's KtxTableListEntry was a 2-level { schema, name }, so
qualifiedTableId always wrote db.name into enabled_tables. When BigQuery,
Snowflake, or SQL Server later ran fast ingest, their introspect step filtered
the scope set with scopedTableNames(scope, { catalog: projectId|database, db })
— catalog was non-null on the introspect side but null in the scope refs, so
every entry was rejected, the live-database adapter staged zero table files,
and detect() failed with 'Adapter "live-database" did not recognize fetched
source output'.

Align the picker boundary with the canonical 3-level KtxTableRef:

- Add catalog: string | null to KtxTableListEntry.
- BigQuery/Snowflake/SQL Server listTables populate catalog from the
  resolved projectId / database; Postgres/MySQL/ClickHouse/SQLite set null.
- qualifiedTableId emits catalog.schema.name when catalog is non-null
  (resolveEnabledTables already accepts the 3-part shape) and
  schemasFromEnabledTables now goes through parseDottedTableEntry so it
  recovers the schema correctly from both 2-part and 3-part entries.
- Export parseDottedTableEntry from enabled-tables.ts (@internal) for picker
  reuse.

Update listTables expectations in all seven connector tests and the setup /
picker test fixtures. Add a picker regression test that covers the
catalog-bearing round-trip (save + refine).

* fix(cli): allow debug telemetry under opt-out env

docs/code-design.md

@@ -89,3 +89,41 @@ enough reason to fix it even when the local code "works."
   (`loadX` vs `loadHigherX`, `createY` vs `createDefaultY`, `xClient`
   vs `xService`), assume callers will pick the wrong one. Unify, or
   document inline why both must exist.
+
+## Dispatch and contract leaks across per-variant layers
+
+Layers with multiple per-variant implementations (warehouse drivers,
+dialects, LLM providers, ingest adapters, historic-SQL probes) drift
+toward parallel switches and informal contracts. The patterns below
+look locally reasonable per file but multiply with the number of
+variants times the number of consumers — every fix has to be applied
+N times, and silent drift between variants is invisible until a user
+hits it.
+
+- **MUST NOT**: Maintain two or more files that switch on the same
+  enum or string union to dispatch to per-variant behavior. Promote
+  the dispatch to a single registry table keyed by the union, exposed
+  through one resolution function. If you find yourself writing the
+  third such switch, the second one was already a bug.
+- **MUST**: When every variant of an abstraction implements the same
+  method, the method belongs on the shared interface. An informal
+  contract that every implementation happens to satisfy is a leak
+  waiting to happen — callers will reach for the concrete class
+  instead of the contract, and the next variant added will silently
+  forget to implement it.
+- **MUST**: When a layer has both a thin shared interface and rich
+  per-variant concrete classes, they must agree. Either widen the
+  interface so callers never need the concrete class, or make the
+  concrete class private (test-only `/** @internal */` JSDoc plus a
+  boundary check in `scripts/check-boundaries.mjs`). A class that is
+  public AND has methods the interface does not expose is the exact
+  configuration that produces leaks.
+
+The warehouse driver / dialect layer in
+`packages/cli/src/connectors/<driver>/` plus
+`packages/cli/src/context/connections/{dialects,drivers}.ts` is the
+canonical worked example: per-driver dialect classes carry
+`/** @internal */`, `scripts/check-boundaries.mjs` enforces the import
+boundary, and dispatch lives in the two registry files. Apply the
+same shape to any other per-variant layer that grows beyond two
+implementations.