feat(execute): allow up to 15 recommendations

Lift stage 3 cap via MAX_RECOMMENDATIONS so UI can render more actions.

Add prompt placeholder replacement and balance deploy vs move logic.

Refresh HUD copy and prompt docs with new tooling and tick pacing.

README.md

@@ -70,7 +70,7 @@ tags:
 
 1. **Assess (Stage 1)** – Calls the analytic MCP tools to classify intensity, coverage, building threats, idle units, and threat level.
 2. **Plan (Stage 2)** – Chooses a strategy (`deploy_new`, `optimize_existing`, `balanced`, etc.) and determines how many actions are justified.
-3. **Execute (Stage 3)** – Emits up to four JSON recommendations (deploy/move/remove) that the service can auto-execute or queue for the player.
+3. **Execute (Stage 3)** – Emits JSON recommendations (deploy/move/remove) that the service can auto-execute or queue for the player.
 4. **Cycle Summary (Stage 4)** – Condenses every loop into a headline, highlights, risks, and next-focus bullets for the Gradio timeline.
 5. **After-Action Report** – Once the scenario ends, the agent merges all summaries + metrics into a “battle report” overlay with charts and actionable follow-ups.
 6. **Human-in-the-loop** – Players can pause, inspect reasoning, toggle auto-execute, or manually override/augment deployments at any point.
@@ -82,7 +82,7 @@ Prompts for every stage live in `prompts.yaml`, making it easy to retune instruc
 ## 🎮 Gameplay Loop & UI Experience
 
 - Start/reset from the control bar or open the accordion to tweak **fire count**, **intensity**, **building cluster size**, **max unit slots**, and **seed**.
-- The advisor refreshes roughly every 10 ticks (≈10 seconds). When it is “thinking,” the HUD animates, and the chat timeline streams stage-by-stage logs.
+- The advisor refreshes roughly every 10 ticks. When it is “thinking,” the HUD animates, and the chat timeline streams stage-by-stage logs.
 - **Auto-Execute** default = ON. Turn it off to require manual approvals (or to stress-test AI reasoning while you handle deployments yourself).
 - Click any grid cell to deploy trucks/helis, remove an existing unit, or even **ignite a new fire** (`🔥 Fire` option) in sandbox mode.
 - Event log + player action chips record everything the human does (deploy, remove, ignite) for inclusion inside the after-action report.

agent.py

@@ -48,9 +48,13 @@ def get_hf_token() -> str | None:
 
 
 # =============================================================================
-# Load Prompts from Configuration
+# Stage 3 action cap + Prompt loading
 # =============================================================================
 
+# Stage 3 UI uses MAX_RECOMMENDATIONS to control how many actions can be rendered
+MAX_RECOMMENDATIONS = 15
+PROMPT_PLACEHOLDERS = {"{{MAX_RECOMMENDATIONS}}": str(MAX_RECOMMENDATIONS)}
+
 def load_prompts() -> dict:
     """Load prompts from prompts.yaml configuration file."""
     prompts_path = Path(__file__).parent / "prompts.yaml"
@@ -59,7 +63,19 @@ def load_prompts() -> dict:
             return yaml.safe_load(f)
     return {}
 
-PROMPTS_CONFIG = load_prompts()
+def _apply_prompt_placeholders(value):
+    """Recursively replace placeholder tokens inside prompt config."""
+    if isinstance(value, str):
+        for token, replacement in PROMPT_PLACEHOLDERS.items():
+            value = value.replace(token, replacement)
+        return value
+    if isinstance(value, dict):
+        return {k: _apply_prompt_placeholders(v) for k, v in value.items()}
+    if isinstance(value, list):
+        return [_apply_prompt_placeholders(item) for item in value]
+    return value
+
+PROMPTS_CONFIG = _apply_prompt_placeholders(load_prompts())
 
 
 # =============================================================================
@@ -757,7 +773,7 @@ Building positions: {json.dumps([(b["x"], b["y"]) for b in buildings[:15]])}
 INSTRUCTIONS:
 1. FIRST generate MOVE actions for ineffective units → move them to uncovered fires
 2. THEN generate DEPLOY actions if more units needed
-3. Max 4 recommendations total
+3. Max {MAX_RECOMMENDATIONS} recommendations total
 4. Remember: deploy ADJACENT to fire (1-2 cells away), not ON the fire
 
 Output format:
@@ -779,7 +795,7 @@ Output format:
             building_positions = set((b["x"], b["y"]) for b in buildings)
             used_positions = set()
             
-            for rec in raw_recs[:4]:  # Limit to 4
+            for rec in raw_recs[:MAX_RECOMMENDATIONS]:  # Limit to UI capacity
                 action = rec.get("action", "deploy")
                 unit_type = rec.get("unit_type", "fire_truck")
                 target = rec.get("target", {})
@@ -1032,7 +1048,74 @@ OUTPUT FORMAT:
                     action="deploy"
                 ))
         
-        return recommendations[:4]  # Cap at 4 for UI display
+        return self._prioritize_recommendations(
+            recommendations,
+            plan.deploy_count,
+            smart_deploy_count,
+        )
+
+    def _prioritize_recommendations(
+        self,
+        recommendations: list[Recommendation],
+        plan_deploy_target: int,
+        smart_deploy_target: int,
+        max_actions: int = MAX_RECOMMENDATIONS,
+    ) -> list[Recommendation]:
+        """
+        Ensure we return a balanced mix of move/deploy actions without exceeding UI limits.
+        """
+        if len(recommendations) <= max_actions:
+            return recommendations
+        
+        deploy_recs = [rec for rec in recommendations if rec.action == "deploy"]
+        move_recs = [rec for rec in recommendations if rec.action == "move"]
+        other_recs = [rec for rec in recommendations if rec.action not in ("deploy", "move")]
+        
+        deploy_priority = 0
+        if deploy_recs:
+            deploy_priority = max(plan_deploy_target, smart_deploy_target, 1)
+        move_priority = len(move_recs)
+        other_priority = len(other_recs)
+        
+        priority_pairs = []
+        if deploy_priority:
+            priority_pairs.append(("deploy", deploy_priority))
+        if move_priority:
+            priority_pairs.append(("move", move_priority))
+        if other_priority:
+            priority_pairs.append(("other", other_priority))
+        
+        if not priority_pairs:
+            return recommendations[:max_actions]
+        
+        priority_pairs.sort(key=lambda item: item[1], reverse=True)
+        ordered_types = [ptype for ptype, _ in priority_pairs]
+        
+        # Ensure every action type gets a chance once primary priorities are exhausted
+        for action_type in ("deploy", "move", "other"):
+            if action_type not in ordered_types:
+                ordered_types.append(action_type)
+        
+        pools = {"deploy": deploy_recs, "move": move_recs, "other": other_recs}
+        indices = {key: 0 for key in pools}
+        selected: list[Recommendation] = []
+        
+        while len(selected) < max_actions:
+            added = False
+            for action_type in ordered_types:
+                pool = pools[action_type]
+                idx = indices[action_type]
+                if idx >= len(pool):
+                    continue
+                selected.append(pool[idx])
+                indices[action_type] += 1
+                added = True
+                if len(selected) >= max_actions:
+                    break
+            if not added:
+                break
+        
+        return selected
 
     # =========================================================================
     # After-Action Report

app.py

@@ -1632,7 +1632,7 @@ def create_app() -> gr.Blocks:
   - 🚁 **Helicopter:** Wide coverage (25%), covers 2 tiles outward from its center — best for large-area control
 - **Settings & Controls:** Use the panel below to quickly tune scenario difficulty (fires, buildings, units, randomness) before sending the team in
 
-**🏆 Win:** Extinguish all fires | **💀 Lose:** Building ≤ 50%
+**🏆 Win:** Extinguish all fires | **💀 Lose:** Building ≤ 50% or Tick ≥ 200 (time out)
         """, elem_classes=["how-to-play"])
         
         # Collapsible Controls Section
@@ -1640,7 +1640,7 @@ def create_app() -> gr.Blocks:
             with gr.Row():
                 with gr.Column(scale=1):
                     fire_count = gr.Slider(
-                        minimum=1, maximum=30, value=15, step=1,
+                        minimum=1, maximum=40, value=20, step=1,
                         label="🔥 Initial Fire Count",
                         info="Number of fire starting points (1-25)"
                     )
@@ -1688,14 +1688,8 @@ def create_app() -> gr.Blocks:
             with gr.Column(scale=2, min_width=300):
                 service = get_service()
                 advisor_interval_ticks = getattr(service, "advisor_interval", 10)
-                tick_interval_seconds = getattr(service, "tick_interval", 1.0)
-                advisor_interval_seconds = advisor_interval_ticks * tick_interval_seconds
-                if isinstance(advisor_interval_seconds, float) and advisor_interval_seconds.is_integer():
-                    advisor_interval_display = int(advisor_interval_seconds)
-                else:
-                    advisor_interval_display = round(advisor_interval_seconds, 1)
                 gr.Markdown(
-                    f"## 🤖 AI Tactical Advisor · (refreshes ~every {advisor_interval_display} seconds)"
+                    f"## 🤖 AI Tactical Advisor · (refreshes every {advisor_interval_ticks} ticks)"
                 )
                 auto_execute_toggle = gr.Checkbox(
                     label="🎮 Auto-Execute",

prompts.yaml

@@ -106,6 +106,8 @@ plan:
     - `find_uncovered_fires()` → Fires needing coverage
     - `find_idle_units()` → Units to reposition first
     - `analyze_coverage()` → Full tactical view
+    - `remove_unit(x, y)` → Free useless/blocked units so slots can be redeployed
+    - `deploy_unit(unit_type, x, y)` → Insert new trucks/helis exactly where the plan decides
     
     STRATEGIC PRIORITIES (STRICT ORDER):
     
@@ -264,7 +266,6 @@ execute:
     RECOMMENDATION COUNT LOGIC:
     - recommendations = reposition_count + deploy_count
     - If uncovered building threats exist: add extra for EACH uncovered threat
-    - Cap at 4 for UI display
     
     CRITICAL RULES:
     1. Deploy 1-2 cells ADJACENT to fire (not ON the fire)