a8b573f9

By: ValentinHirschi <valentin.hirschi@gmail.com>

enable amplitude joint sampling with retained proposal policies

Bind routed shared-energy intersections through the existing amplitude map
catalogue. Prepare each draw's discrete compact/fallback and radius choices at
fixed Arb precision before stability selection, then require complete native
agreement through selected, summed, direct, debug and retry paths. Native
geometry and Jacobians are reconstructed in each precision.

Cover empty prerequisites, foreign exact/proxy partitions, actual body-triggered
rescue and all-orientation two-loop physical/reference evaluation. Keep hosted
cross-section joints gated on authoritative cut data. Record the user's sampling
runtime budget and the remaining GL638 geometry and performance gates.

Validation: 170 focused core tests plus the generated kite acceptance pass
(8192 joint-reference draws, normalization/raw moments, all 18 physical
orientations and original-source retry). Both saved-state/canonical-channel
API regressions pass. Core/API cargo check, formatting, diff checks and
changed-line clippy checks pass. New proposal records are transient metadata;
no serialized state fields are added.

User prompts:

Just a status question—please continue your current orchestration unchanged: where does the joint two-normal density sit in the implementation sequence, and what prerequisites remain before you reach it? (direct-H focusing alone is not expected to bound the previously identified corner’s weights.)

(continue as planned but make sure the time spent in the sampling maps is still quite significantly less than the actual GL638 all-orientation integrand, i.e. it's not crucial to optimize the runtime of the sampling, except if it is somehow comparable to the integrand time; it should be like 10% at the very most, once below that then no need to optimize further).

Timeline