Task: Construct every bond product by name

Every bond code that routes to the bond instrument table has a named constructor, and each constructor states the facts that product has.

• Each of the ten routed bond codes has a named constructor, and each refuses what its product cannot state.
• A bond product is appended to the instrument batch by one seam, not by nine hand-wired vectors.
• The constructors have tests that state what each product carries and what it refuses.

(Implementation strategy. Written when work starts; key decisions are distilled into the parent story's * Decisions at close, but the plan itself stays — it is the historical record of what we did.)

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.

Ten constructors, one per routed bond code, and a single append that crosses a whole product into the instrument batch. Each constructor states only what its product has, and each refuses what its product cannot be: an option with no type, a forward with no settlement date, a future with no contract, a repo with neither rate nor index, a total return swap with neither return nor funding leg, a callable bond with no call date, a convertible with no target.

The batch the constructors fill is generated from the trade type catalogue's `* Batch/Children` list, and the bond fact rows were missing from it. A first draft of this note said the omission had broken main. It had not: the security's issue rows travel on the plan's slot, which is hand-written, and the arrays the regeneration deleted from the batch were never read. What was true is narrower and still worth fixing — the model now states the bond children, so a regeneration keeps the arrays the products need, and the four fact rows the constructors fill exist in the batch rather than only in a hand-added header.

Tests: thirteen cases in `domain_bond_instruments_tests.cpp`, asserting each product's own row, the refusals above, and that a product crosses the batch in one append. `ores.trading.api.tests "[bond]"` passes with 51 assertions.

The full tree was building when this was written and the build broke elsewhere; the whole-tree run is owed once that is fixed.

The whole tree is green: `./compass.sh build` finished, `ores.trading.api.tests` passes with 618 assertions in 162 cases, and `ores.synthetic.core.tests` with 2,050 in 82. The earlier note that the build was broken elsewhere no longer applies: main was repaired and this branch was rebased onto it.