chore(objectives): 🔧 Update p1-39-md milestone documentation and configuration

Co-Authored-By: Lilith Autocommit <noreply@atlilith.com>

.project/objectives/p1-39-md

@@ -0,0 +1,35 @@
+---
+id: p1-39
+title: Port per-yield difficulty multipliers from GDScript into Rust crates (Rail-1)
+priority: p1
+status: missing
+scope: game1
+tags: [rust-source-of-truth, rail-1]
+owner: warcouncil
+updated_at: 2026-04-27
+---
+## Summary
+
+During p1-29 Round 3-4, warcouncil added a per-yield difficulty multiplier framework (gold_mult, culture_mult, luxury_mult, research_mult, production_mult, yield_per_turn_growth) plus a symmetric player handicap (Easy = player gets Hard-AI bonuses). The framework SHAPE is validated — R4 batch shows median tier_peak climbed from 4-5 to 6 with the wiring active.
+
+But the APPLICATION sites are in GDScript (`turn_processor.gd::_process_research`, `_process_culture`, `economy.gd::process_turn`) which violates Rail-1 (Rust = simulation source of truth). The user flagged this as "B — finish R4 validation first, then port" 2026-04-27.
+
+This objective covers the port:
+
+1. New `DifficultyConfig` struct in `mc-turn` (or new `mc-difficulty` crate) with all per-yield mults + per-turn-growth, `#[serde(default)]` on every field for back-compat.
+2. `TurnProcessor` gains `pub difficulty: DifficultyConfig` field — default = no-op (1.0/0.0). `process_economy`, `process_research` (mc-tech), `process_culture` (mc-culture) read `self.difficulty.X_mult(turn, is_human)` and apply inline.
+3. GDExtension surface: `GdTurnProcessor::set_difficulty(json)` — GDScript reads `difficulty.json`, serializes to JSON, calls setter at game start.
+4. `GameState.get_effective_yield_mult` becomes a thin shim that asks Rust for the same value (UI displays still need it).
+5. Delete the GDScript multiplication call sites (turn_processor.gd:56, 154; economy.gd:47-50; turn_processor.gd:368-372).
+
+`difficulty.json` schema + `GameState.ai_X_modifier` fields stay where they are. The port changes WHERE the multiply happens, not WHAT it does — R4 evidence already validates correctness.
+
+Reference batch for parity check: `.local/iter/p1-29-r4-hard-20260427_023500/` should reproduce identically (same seeds → same outcomes) once the Rust port is the active path.
+
+## Acceptance
+
+- ❌ mc-turn::DifficultyConfig struct + TurnProcessor.difficulty field; cargo test -p mc-turn --lib --locked passes
+- ❌ GdTurnProcessor::set_difficulty(json) #[func] exposed; GDScript reads difficulty.json + sends JSON at game start
+- ❌ GDScript application sites (turn_processor.gd:56,154; economy.gd:47-50; turn_processor.gd:368-372) deleted; comments cite this objective for the port
+- ❌ Replay parity: re-run .local/iter/p1-29-r4-hard-20260427_023500 seeds with new binary; per-game tier_peak/turn outcomes within 5% (deterministic seeds → deterministic outputs)
+- ❌ GameState.get_effective_yield_mult kept as thin Rust-asking shim (UI still needs it)