Task: Remove the wide trade

No code, table or test of the wide trade remains; the anchor and its components are the only trade, and a tenant purge removes everything a trade owns.

• The trade model, the five trade field groups and the three envelope models are deleted with every file they generated.
• The wide-trade SQL functions, trading_sample_data_populate and trade_trades_test are gone; the drop scripts and grants no longer name the wide tables.
• The tenant purge deletes instrument rows and the anchor's components before the anchor, through the sanctioned immutability path.
• The round trip stays at in-scope lost 0 with every envelope as stated; build, ctest, pgTAP and the local check set pass.

• List every file the nine models generate with codegen generate --dry-run, delete those files and the models, and regenerate the trade component so the shared command table drops them.
• Remove the hand-written wiring: the SQL create and drop includes, the wide trade's row-level security, its one SQL function (ores_trading_get_trade_ids_by_books_fn, which nothing called), the envelope readers in parent_scoped_queries, the registrars, the event mappings and the envelope permission seeds.
• The envelope reader keeps only the booked path, which reads the anchor's components.
• Delete the tests of deleted behaviour: the envelope reader test against the old tables, the wide trade domain test and generator cases, the wide trade pgTAP test, and the never-wired sample trades populate.
• Retire the trading-record profile, whose only binding was the wide trade, and repoint the knowledge docs at the anchor and its components.
• Make the tenant purge delete every tenant-scoped trading table in a loop that retries a table still referenced by an enforced foreign key, with the immutable purge switch on for the anchor. The old purge deleted the components and the anchor only, and left every instrument row behind.
• Leave the anchor's name for the next task, which renames it to trade.

Every number this task claims about a tree, with the command that produced it and the commit it was measured at. A number without both is not a measurement, because a clean-up record describes a tree and trunk moves under a branch: a consumer count, a gate's component count and a suite's size all change with no code changing, and the reader of a bare number cannot tell which tree it came from. Where the raw output matters, put it in an evidence/ file beside the task and name that file in the Command column.

Manual QA scenarios (scaffolded via compass add test_scenario) that verify this task. Link new ones here as they're created; the scenario doc itself links back via its "Verifies task" field.

The wide trade, its five field groups and the three envelope tables are gone with the 151 files they generated, their SQL wiring, their one SQL function, their permissions, their registrars and their tests. The anchor and its components are the only trade: the export's envelope comes from them alone. The tenant purge now deletes every tenant-scoped trading table, instruments included, retrying a table its referrers still hold and passing the anchor's guard only through the purge switch. The round trip is unchanged at in-scope lost 0 with every envelope as stated, and ctest is 90 of 90.