CRITICAL RUNTIME FIX: Remove tuple unpacking from environment handlers - graders now return FLOAT only, not tuples

server/config_debug_environment.py

@@ -56,15 +56,25 @@ class ConfigDebugEnvironment(Environment):
         task_id = self._current_task_id()
         task = get_task(task_id)
 
-        # Run the grader
-        reward, error_message, bugs_fixed = task.grader(action.fixed_config)
-        reward = max(0.0, min(1.0, reward))
+        # Run the grader - now returns FLOAT ONLY for validator compatibility
+        grader_result = task.grader(action.fixed_config)
+        
+        # Handle result gracefully (should be float)
+        if isinstance(grader_result, float):
+            reward = grader_result
+        elif isinstance(grader_result, tuple) and len(grader_result) > 0:
+            # Fallback for legacy tuple format
+            reward = grader_result[0]
+        else:
+            reward = 0.01  # Safe default
+        
+        reward = max(0.01, min(0.99, reward))
 
         self.current_step += 1
         self._global_step += 1
-        self.bugs_found_so_far = len(bugs_fixed)
+        self.bugs_found_so_far = 0  # Default since grader no longer returns this
         self.previous_reward = round(reward, 4)
-        self.current_error_message = error_message
+        self.current_error_message = ""  # Default since grader no longer returns this
 
         # Check if task is complete
         task_done = reward >= 0.99 or self.current_step >= MAX_STEPS_PER_TASK

server/env.py

@@ -284,13 +284,25 @@ def ui_step(fixed_config):
 
     task_id = _get_current_task_id()
     task = get_task(task_id)
-    reward, error_message, bugs_fixed = task.grader(fixed_config)
+    
+    # Run the grader - now returns FLOAT ONLY for validator compatibility
+    grader_result = task.grader(fixed_config)
+    
+    # Handle result gracefully (should be float)
+    if isinstance(grader_result, float):
+        reward = grader_result
+    elif isinstance(grader_result, tuple) and len(grader_result) > 0:
+        # Fallback for legacy tuple format
+        reward = grader_result[0]
+    else:
+        reward = 0.01  # Safe default
+    
     reward = max(0.01, min(0.99, reward))
 
     env_state.current_step += 1
-    env_state.bugs_found_so_far = len(bugs_fixed)
+    env_state.bugs_found_so_far = 0  # Default since grader no longer returns this
     env_state.previous_reward = round(reward, 4)
-    env_state.current_error_message = error_message
+    env_state.current_error_message = ""  # Default since grader no longer returns this
 
     task_done = reward >= 0.99 or env_state.current_step >= MAX_STEPS_PER_TASK