feat: Add mcrfpy.step() and synchronous screenshot for headless mode (closes #153)

Implements Python-controlled simulation advancement for headless mode:

- Add mcrfpy.step(dt) to advance simulation by dt seconds
- step(None) advances to next scheduled event (timer/animation)
- Timers use simulation_time in headless mode for deterministic behavior
- automation.screenshot() now renders synchronously in headless mode
  (captures current state, not previous frame)

This enables LLM agent orchestration (#156) by allowing:
- Set perspective, take screenshot, query LLM - all synchronous
- Deterministic simulation control without frame timing issues
- Event-driven advancement with step(None)

🤖 Generated with [Claude Code](https://claude.com/claude-code)

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

src/GameEngine.h

@@ -110,6 +110,10 @@ private:
     bool headless = false;
     McRogueFaceConfig config;
     bool cleaned_up = false;
+
+    // #153 - Headless simulation control
+    int simulation_time = 0;           // Simulated time in milliseconds (for headless mode)
+    bool simulation_clock_paused = false;  // True when simulation is paused (waiting for step())
     
     // Window state tracking
     bool vsync_enabled = false;
@@ -189,6 +193,11 @@ public:
     std::string getViewportModeString() const;
     sf::Vector2f windowToGameCoords(const sf::Vector2f& windowPos) const;
 
+    // #153 - Headless simulation control
+    float step(float dt = -1.0f);  // Advance simulation; dt<0 means advance to next event
+    int getSimulationTime() const { return simulation_time; }
+    void renderScene();  // Force render current scene (for synchronous screenshot)
+
     // global textures for scripts to access
     std::vector<IndexTexture> textures;