Round 2 Implementation: Multi-period campaign, regulatory shocks, adversarial grading, and GRPO training infrastructure

Refine comment wording for scoring stability paths

Harden score constraints in app and align inference Round-1 defaults

Harden inference score/reward logging for validator

Scrub remaining instances of raw 0.0 and 1.0 logic to 0.01 and 0.99

Clamp grader and inference rewards securely to 0.01 and 0.99

Fix inference stdout formatting and update graders tests for strictly clamped reward constraint bounds

fix: global invariant enforcement 0<score<1 on api pit stop and robust tests

fix: increase score epsilon to 0.01 for validator safety

fix: clamp scores to (0,1) exclusive for Phase 2 validation; add root requirements.txt; defensive imports in inference.py

parameter compliance

parameter compliance

docs: update baseline scores and add failure analysis

fix: resolve README merge conflicts + add mandatory [START]/[STEP]/[END] structured logging to inference.py

Added hackathon-compliant inference.py and validation script