Overhaul model system with multi-stage training, variants, benchmarks, and eval

Replace the single-stage training + flat capability score with a realistic AI
development pipeline: pre-training with Chinchilla scaling laws, SFT with
specializations, alignment with safety/capability tradeoffs (RLHF/DPO/Constitutional),
model families with distillation/fine-tuning/quantization variants, named benchmark
suite with compute-costing eval jobs, and segment-specific market quality.

Phases 1-6 of the model rework plan: new types, engine rewrite, save migration,
training events/risk system, concurrent training, variant creation, benchmark
evaluation with leaderboard, and market integration.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

packages/game-engine/src/systems/achievementSystem.ts

@@ -9,7 +9,7 @@ const ERA_INDEX: Record<string, number> = { startup: 0, scaleup: 1, bigtech: 2,
 
 function getFieldValue(state: GameState, field: string): number {
   if (field === 'meta._eraIndex') return ERA_INDEX[state.meta.currentEra] ?? 0;
-  if (field === 'meta._deployedModelCount') return state.models.trainedModels.filter(m => m.isDeployed).length;
+  if (field === 'meta._deployedModelCount') return state.models.baseModels.filter(m => m.isDeployed).length;
   const parts = field.split('.');
   let current: unknown = state;
   for (const part of parts) {