Skip to content

Validation strategy for ASRQuant 1.2.0

Verified locally

  • 74 research, numerical-modelling, backtest, statistics and visualization tests pass.
  • 62 production-readiness and guarded-live tests pass.
  • Aggregate automated tests: 136.
  • Production-boundary line coverage: 93% (production.py, audit_store.py, live.py).
  • Live authorization fails closed without a valid certificate and explicit environment arm.
  • Audit hash-chain tampering, duplicate order identifiers, stale data, price collars, capital limits, reconciliation mismatches and persistent kill-switch behavior are tested.

Evidence that must come from CI or the target operator

  • supported OS/Python matrix;
  • static analysis and type-check results;
  • vulnerability, dependency and secret scans;
  • full transitive SBOM and artifact attestation;
  • reproducible-build comparison;
  • real broker-paper longevity and order-state evidence;
  • monitoring/alert delivery, backup/restore, disaster recovery and rollback exercises;
  • legal/compliance, data-license and independent model-validation approvals.

The local validation report is VALIDATION_v1.2.0.md. A live deployment must additionally produce a passing ProductionReadinessReport for its own environment.