Add env.step safety net so checkpoint hallucinations never crash evaluation.

Harden checkpoint sanitizer against all sandbox/sim edge cases.

Finalize checkpoint-first evaluation and sandbox submission workflow.

Improve checkpoint policy quality and environment robustness.

Fix checkpoint evaluation reliability and keep heuristic as strict fallback.

audit pass: hackathon-ready polish and truthful reporting

Polish judge-facing docs and add checkpoint evaluation.

Harden training signal and clean generated artifacts.

Harden hackathon readiness and surface reproducible evidence.

updated EICC v2 environment, APIs, and training pipeline