test + docs: lock should_execute_tools guard semantics (#3220)

Two small follow-ups to the guard:

1. Fix the should_execute_tools docstring so it matches the actual code.
   The previous version said "Only execute when finish_reason explicitly
   signals tool intent" but the code also accepts finish_reason == "stop".
   Explain why (some compliant providers emit "stop" with legitimate tool
   calls — openai_compat_provider.py already mirrors this at lines ~633 /
   ~678 where ("tool_calls", "stop") are both treated as the terminal
   tool-call state). Without this, a strict "tool_calls"-only guard would
   regress 15 existing runner tests that construct LLMResponse with
   tool_calls but no explicit finish_reason (default = "stop").

2. Add tests/providers/test_llm_response.py. This locks the three cases:
   - no tool calls                  -> never executes
   - tool calls + "tool_calls"/stop -> executes
   - tool calls + refusal / content_filter / error / length / ... -> blocked

   These are exactly the boundary cases the #3220 fix is about; without a
   test here a future refactor could silently revert the guard.

Body + tests only, no behavior change beyond the existing PR's intent.

Made-with: Cursor

nanobot/providers/base.py

@@ -75,17 +75,18 @@ class LLMResponse:
     def should_execute_tools(self) -> bool:
         """Check if tool calls should be executed (guards against gateway injection).
 
-        Only execute when finish_reason explicitly signals tool intent.
-        Tool calls under any other finish_reason (refusal, content_filter, error, etc.)
-        are treated as anomalous and should not be executed.
+        Executes only when ``has_tool_calls`` is true and ``finish_reason`` is one of
+        the known-good signals: ``"tool_calls"`` (explicit intent) or ``"stop"`` (some
+        compliant providers emit ``stop`` for legitimate tool calls; existing paths in
+        ``openai_compat_provider`` already treat both as the tool-call terminal state).
+
+        Tool calls under any other ``finish_reason`` (e.g. ``refusal``, ``content_filter``,
+        ``error``) are treated as anomalous — typically injected by non-compliant API
+        gateways — and are skipped.
         """
         if not self.has_tool_calls:
             return False
-        if self.finish_reason == "tool_calls":
-            return True
-        if self.finish_reason == "stop":
-            return True
-        return False
+        return self.finish_reason in ("tool_calls", "stop")
 
 
 @dataclass(frozen=True)