Skip to content

Safety Scorecard

Multi-dimensional safety assessment that orchestrates simulation, threat analysis, Monte Carlo risk analysis, and policy evaluation to produce a comprehensive scorecard with letter grades (A+ through F) per dimension.

Quick Start

from replication.scorecard import SafetyScorecard, ScorecardConfig

sc = SafetyScorecard()
result = sc.evaluate()
print(result.render())
print(f"Overall: {result.overall_grade} ({result.overall_score}/100)")

Key Classes

  • SafetyScorecard — Runs multiple analysis passes (simulation, threats, Monte Carlo, policy) and combines results into a single graded scorecard.
  • ScorecardResult — Contains per-dimension scores, overall grade, detailed breakdown, and rendered report.
  • DimensionScore — Score for a single safety dimension (0–100) with letter grade and contributing metrics.
  • ScorecardConfig — Configuration: scenario preset, strategy, max depth, Monte Carlo run count, policy preset.

Grading Scale

Grade Score
A+ 97–100
A 93–96
A− 90–92
B+ 87–89
B 83–86
… …
F 0–59

CLI

python -m replication.scorecard                           # default
python -m replication.scorecard --scenario balanced       # from preset
python -m replication.scorecard --mc-runs 50              # Monte Carlo runs
python -m replication.scorecard --policy strict           # policy preset
python -m replication.scorecard --json                    # JSON output
python -m replication.scorecard --quick                   # skip slow analyses