# Delta Verdict Engine

## Module
Policy

## Status
IMPLEMENTED

## Description
Full delta verdict computation comparing two evaluation states, with signed delta JSON, API endpoints for delta generation, and verdict ID generation.

## Implementation Details
- **WhatIfSimulationService**: `src/Policy/StellaOps.Policy.Engine/WhatIfSimulation/WhatIfSimulationService.cs` -- `WhatIfSimulationService` (internal sealed class)
  - `SimulateAsync(WhatIfSimulationRequest)` computes delta between baseline and simulated evaluation states
  - Supports SBOM diffs: add, remove, upgrade, downgrade operations
  - Computes `WhatIfDecisionChange`: status_changed, severity_changed, new, removed
  - Decision simulation: new components checked against advisory count, VEX override, reachability downgrade
  - Upgrade simulation: fixed-all -> allow, remaining advisories -> warn
  - Downgrade simulation: with advisories -> deny (higher priority 150)
  - `WhatIfSummary`: TotalEvaluated, TotalChanged, NewlyAffected, NoLongerAffected, StatusChanges, SeverityChanges, Impact
  - `WhatIfImpact`: risk delta (increased/decreased/unchanged), blocked/warning deltas, recommendation text
  - Simulation ID generation: `whatif-{SHA256(seed)[..16]}`
- **WhatIfSimulationModels**: `src/Policy/StellaOps.Policy.Engine/WhatIfSimulation/WhatIfSimulationModels.cs` -- request/response DTOs
  - `WhatIfSimulationRequest`: TenantId, BaseSnapshotId, SbomDiffs, DraftPolicy, TargetPurls, IncludeExplanations, Limit, CorrelationId
  - `WhatIfSbomDiff`: Purl, Operation, OriginalVersion, NewVersion, AdvisoryIds, VexStatus, Reachability
  - `WhatIfDecision`: Status, Severity, RuleName, Priority, HasException
  - `WhatIfExplanation`: MatchedRules, Factors, VexEvidence, Reachability
- **ConsoleSimulationDiffService**: `src/Policy/StellaOps.Policy.Engine/Console/ConsoleSimulationDiffService.cs` -- deterministic delta diff for console surface
  - Before/After severity breakdowns, delta (added/removed/regressed), rule impact, explain samples
- **SimulationAnalyticsService**: `src/Policy/StellaOps.Policy.Engine/Simulation/SimulationAnalyticsService.cs` -- delta summary computation with severity change tracking
- **Effective Decision Map**: `src/Policy/StellaOps.Policy.Engine/EffectiveDecisionMap/` -- materialized baseline decisions for delta comparison
- **Simulation Endpoints**: `src/Policy/StellaOps.Policy.Engine/Endpoints/`
  - `ConsoleSimulationEndpoint.cs` -- console surface simulation diff
  - `OverlaySimulationEndpoint.cs` -- overlay-based simulation
  - `RiskSimulationEndpoints.cs` -- risk simulation with breakdowns
- **Attestation**: `src/Policy/StellaOps.Policy.Engine/Attestation/` -- verdict attestation for signed delta output

## E2E Test Plan
- [ ] POST what-if simulation with add component (3 advisories); verify deny decision with severity=high
- [ ] POST what-if simulation with remove component; verify decision=allow and change_type=removed
- [ ] POST what-if simulation with upgrade component fixing all CVEs; verify decision=allow
- [ ] POST what-if simulation with downgrade component with advisories; verify decision=deny with priority 150
- [ ] POST what-if simulation with VEX not_affected override; verify deny overridden to allow
- [ ] POST what-if simulation with unreachable finding; verify deny downgraded to warn
- [ ] Verify delta summary: TotalChanged matches actual number of decision changes
- [ ] Verify impact: risk delta is "increased" when blocked count goes up, "decreased" when it goes down
- [ ] POST with IncludeExplanations=true; verify explanations contain matched rules, SBOM factors, VEX evidence, and reachability
- [ ] POST console simulation diff with two policy versions; verify deterministic before/after severity breakdowns and delta counts