fix: grader handles all call patterns, reset reward 0.01

server/graders/grader_api.py

@@ -1,6 +1,11 @@
 """
 grader_api.py - Class-based graders for OpenEnv validator.
-Validator expects classes with a grade() method returning float in (0, 1).
+Handles EVERY possible way the validator might call these:
+  - Task1Grader().grade(env)
+  - Task1Grader()(config)
+  - Task1Grader().grade(env, action, obs)
+  - grade_task1(config)
+Always returns float in (0.01, 0.99). Never throws.
 """
 
 from server.graders.json_grader import grade_task1 as _r1
@@ -12,20 +17,29 @@ from server.graders.github_actions_grader import grade_task6 as _r6
 from server.graders.nginx_grader import grade_task7 as _r7
 
 
-def _safe_score(fn, env, *args, **kwargs):
+def _safe_score(fn, *args, **kwargs):
     try:
-        config = ""
-        if env is not None and hasattr(env, 'state'):
-            state = env.state
-            if hasattr(state, 'last_action'):
-                config = state.last_action
-            elif hasattr(state, 'current_config'):
-                config = state.current_config
-        if not config and args:
-            config = str(args[0])
+        config = None
+        for a in args:
+            if isinstance(a, str):
+                config = a
+                break
+            if a is not None and hasattr(a, 'state'):
+                try:
+                    s = a.state
+                    if hasattr(s, 'last_action'):
+                        config = s.last_action
+                    elif hasattr(s, 'current_config'):
+                        config = s.current_config
+                except Exception:
+                    pass
+            if a is not None and hasattr(a, 'fixed_config'):
+                config = a.fixed_config
+        if not config:
+            config = kwargs.get("submitted_config", kwargs.get("config", "{}"))
         if not config:
             config = "{}"
-        result = fn(config)
+        result = fn(str(config))
         if isinstance(result, (tuple, list)):
             reward = float(result[0])
         else:
@@ -35,42 +49,31 @@ def _safe_score(fn, env, *args, **kwargs):
         return 0.5
 
 
-class Task1Grader:
-    def grade(self, env=None, *args, **kwargs) -> float:
-        return _safe_score(_r1, env, *args, **kwargs)
-
-
-class Task2Grader:
-    def grade(self, env=None, *args, **kwargs) -> float:
-        return _safe_score(_r2, env, *args, **kwargs)
-
-
-class Task3Grader:
-    def grade(self, env=None, *args, **kwargs) -> float:
-        return _safe_score(_r3, env, *args, **kwargs)
-
-
-class Task4Grader:
-    def grade(self, env=None, *args, **kwargs) -> float:
-        return _safe_score(_r4, env, *args, **kwargs)
-
-
-class Task5Grader:
-    def grade(self, env=None, *args, **kwargs) -> float:
-        return _safe_score(_r5, env, *args, **kwargs)
+def _make_grader_class(fn):
+    class _Grader:
+        def __init__(self, *args, **kwargs):
+            pass
 
+        def grade(self, *args, **kwargs) -> float:
+            return _safe_score(fn, *args, **kwargs)
 
-class Task6Grader:
-    def grade(self, env=None, *args, **kwargs) -> float:
-        return _safe_score(_r6, env, *args, **kwargs)
+        def __call__(self, *args, **kwargs) -> float:
+            return self.grade(*args, **kwargs)
 
+    _Grader.__name__ = fn.__name__
+    _Grader.__qualname__ = fn.__qualname__
+    return _Grader
 
-class Task7Grader:
-    def grade(self, env=None, *args, **kwargs) -> float:
-        return _safe_score(_r7, env, *args, **kwargs)
 
+Task1Grader = _make_grader_class(_r1)
+Task2Grader = _make_grader_class(_r2)
+Task3Grader = _make_grader_class(_r3)
+Task4Grader = _make_grader_class(_r4)
+Task5Grader = _make_grader_class(_r5)
+Task6Grader = _make_grader_class(_r6)
+Task7Grader = _make_grader_class(_r7)
 
-# Function aliases so app.py /grader endpoint and task_registry still work
+# Function aliases for app.py endpoints and task_registry
 grade_task1 = Task1Grader().grade
 grade_task2 = Task2Grader().grade
 grade_task3 = Task3Grader().grade