feat: add leakage-safe forex history learning

fix: make paper PnL truthful and gate FinRL-X

perf: bound FinRL-X feature store reads

fix: allow bounded FinRL-X market training

feat: train FinRL-X challengers by market

fix: reserve FinRL-X worker initialization budget

feat: operationalize pinned FinRL-X runner

fix: distinguish completed paper outcomes

feat: add governed FinRL-X quant challenger

fix: separate equity and forex paper evidence

docs: design FinRL-X challenger integration

fix: resume paper trading evidence exploration

feat: publish finance benchmark evaluation

feat: prepare portable benchmark release

feat: release BLUM Finance 4B model

docs: publish BLUM Finance release evidence

feat: add immutable BLUM Finance GPU job

docs: record BLUM Finance dataset release

feat: package guarded BLUM Finance release

feat: gate BLUM Finance promotion on evidence

fix: normalize BLUM Finance training targets

feat: add reproducible BLUM Finance LoRA training

feat: add standalone BLUM Finance client

feat: expose guarded BLUM Finance release export

feat: build leakage-safe BLUM Finance dataset

feat: define BLUM Finance release contracts

docs: plan BLUM Finance model release

docs: design BLUM Finance model release

fix: classify trading ML evidence gaps

fix: bound trading ML learning slices

feat: activate trading ML champion challenger

feat: add purged trading ML validation

fix: make feature store locking portable

fix: serialize trading feature store commits

feat: add incremental trading feature store

fix: persist cost-adjusted equity outcomes

fix: exclude leaked trading ML feature data

fix: harden trading ML feature contracts

feat: add point-in-time trading ML features

feat: add trading ML persistence

docs: plan trading ML champion challenger

docs: define scalable trading ML data plane

docs: design trading ML champion challenger

feat: add hierarchical forex reinforcement

docs: design hierarchical forex reinforcement

fix: restore auditable paper trading runtime

fix: harden runtime health reporting

fix: prioritize actionable equity scouting

feat: add evidence-bound forex reinforcement council

fix: bootstrap forex paper reinforcement learning