Automate failure-driven training continuation

docs/BELGIUM_BUILDING_TRAINING_LOOP.md

@@ -92,6 +92,10 @@ recall are repeated, while true negative train tiles are repeated when a
 regional precision gate or the pure-background gate fails. Calibration, test,
 background-test and validation AOIs are excluded by their frozen corpus split;
 the generated evidence records that no protected sample entered training.
+The checkpointed orchestrator invokes this builder after every rejected
+iteration, stores its checksum in `training-loop-state.json`, and uses the
+resulting dataset YAML for the next checkpoint. A restart resumes both the
+candidate weights and that exact failure-driven training input.
 The orchestrator refuses to start unless every automated frozen-dataset gate
 passes and the corpus contains zero blank/low-variance positive tiles. The
 audit status may remain `needs_human_review` while training and objective

docs/CODEX_EXECUTION_LOG.md

@@ -11773,11 +11773,20 @@ Deployment evidence:
   immutable, automated failures are empty, spatial leakage is `ok`, and blank
   positive-tile count is zero. Human sign-off remains a separate mandatory
   final promotion gate.
+- Closed the next orchestration gap: a rejected iteration now invokes the
+  leak-free failure-driven sampler automatically, records its evidence
+  checksum and next dataset YAML in `training-loop-state.json`, and resumes
+  both the candidate weights and exact sampling input after interruption.
+- Confirmed v37 epoch 1 completed on CUDA with validation precision `0.601`,
+  recall `0.455`, mAP50 `0.474` and mAP50-95 `0.205`; the run remains inactive
+  and these internal-validation metrics are not release evidence.
 
 Verified in this pass:
 
 - `py -3 -m pytest -q backend/tests/test_belgium_training_loop.py backend/tests/test_belgium_training_iteration_assessment.py backend/tests/test_belgium_training_portfolio.py`
   (`12 passed`).
+- `py -3 -m pytest -q backend/tests/test_belgium_training_loop.py backend/tests/test_failure_driven_yolo_sampling.py backend/tests/test_belgium_training_iteration_assessment.py`
+  (`15 passed`).
 
 Open:
 

docs/TODO.md

@@ -972,6 +972,7 @@ This file now starts with the current implementation status. Older preparation/b
 - [x] Reject positive labels over blank/no-data imagery and replace partial SPW 2024 coverage with the complete dated SPW 2023 campaign.
 - [x] Exclude GRB/PICC features created after the corresponding dated imagery period while retaining auditable rejection evidence.
 - [x] Allow the objective CUDA loop to consume an automatically clean `needs_human_review` corpus while keeping final human sign-off as a separate, mandatory promotion gate.
+- [x] Persist checksummed train-only failure-driven sampling after every rejected loop iteration and resume the next checkpoint from that exact dataset YAML.
 - [x] Evaluate the completed v36 YOLO11x checkpoint calibration-first on the rotated v30 holdouts; reject it before opening test/background because the regional calibration gate failed.
 - [ ] Finish and assess the leak-free v37 YOLO11x failure-driven CUDA iteration; open test/background evidence only if every calibration gate passes.