fix(approval): warn on empty smart-approval guardian answers

A 200 response with empty content (finish_reason="length" after a
reasoning model spends the whole max_tokens budget on hidden reasoning)
mapped to ESCALATE via a plain _VERDICTS dict miss with no log above
DEBUG — indistinguishable from a genuine strict verdict for days
(#117428, failure mode of #108163/#68263). The exception path already
warns (#93809); this closes the truncation path's asymmetry.

Fixes #117428

tests/tools/test_smart_approval_injection.py

@@ -166,6 +166,39 @@ class TestSmartApprovePromptHardening(unittest.TestCase):
         mock_call_llm.return_value = self._make_response("I think this is probably fine")
         assert _smart_approve("rm -rf /", "recursive delete") == "escalate"
 
+    @patch("agent.auxiliary_client.call_llm")
+    def test_empty_answer_escalates_with_warning(self, mock_call_llm):
+        """The #117428 failure mode: a 200 response with empty content (finish_reason==
+        "length" after a reasoning model spent the whole max_tokens budget on hidden
+        reasoning) must escalate AND log at WARNING — at default log levels it is otherwise
+        indistinguishable from a genuine ESCALATE verdict."""
+        response = self._make_response("")
+        response.choices[0].finish_reason = "length"
+        mock_call_llm.return_value = response
+        with self.assertLogs("tools.approval", level="WARNING") as logs:
+            assert _smart_approve("rm -rf /", "recursive delete") == "escalate"
+        assert any("empty answer" in message and "length" in message
+                   for message in logs.output), logs.output
+
+    @patch("agent.auxiliary_client.call_llm")
+    def test_empty_answer_without_finish_reason_still_warns(self, mock_call_llm):
+        """A missing finish_reason must not hide the empty-answer WARNING: the field is
+        populated unevenly across OpenAI-compatible providers."""
+        response = self._make_response(None)
+        response.choices[0].finish_reason = None
+        mock_call_llm.return_value = response
+        with self.assertLogs("tools.approval", level="WARNING") as logs:
+            assert _smart_approve("rm -rf /", "recursive delete") == "escalate"
+        assert any("finish_reason=None" in message for message in logs.output), logs.output
+
+    @patch("agent.auxiliary_client.call_llm")
+    def test_recognized_verdict_does_not_warn(self, mock_call_llm):
+        """The WARNING is specific to empty answers — firing on healthy calls too would
+        train operators to skim past it, the exact failure #117428 describes."""
+        mock_call_llm.return_value = self._make_response("APPROVE")
+        with self.assertNoLogs("tools.approval", level="WARNING"):
+            assert _smart_approve("python -c 'print(1)'", "script execution") == "approve"
+
 
 if __name__ == "__main__":
     unittest.main()

tools/approval_smart.py

@@ -112,6 +112,16 @@ def _smart_approve(command: str, description: str) -> str:
         )
         logger.debug("Smart approvals: LLM call completed in %.1fs", time.monotonic() - _smart_t0)
         answer = (response.choices[0].message.content or "").strip().upper()
+        if not answer:
+            # WARNING, not DEBUG: an empty-but-200 body is an infrastructure failure, not a
+            # verdict — typically finish_reason=="length" after a reasoning model spent the
+            # whole max_tokens budget on hidden reasoning (#117428). It escalates like any
+            # uncertain outcome, but is indistinguishable from a genuine ESCALATE in the logs
+            # unless this fires above DEBUG.
+            finish_reason = getattr(response.choices[0], "finish_reason", None)
+            logger.warning("Smart approvals: guardian returned an empty answer "
+                           "(finish_reason=%s), escalating", finish_reason)
+            return "escalate"
         return _VERDICTS.get(answer, "escalate")
     except Exception as e:
         # WARNING, not DEBUG: a failed/blocked guardian call is a real event