Add debug logging to diagnose empty task results

**Problem**: Task "create a gitea repo" completed 86 messages but returned
generic fallback "Task completed (86 messages, $0.71)" with no actual work
done. Zero tool calls tracked, zero assistant messages captured.

**Debug additions**:
1. Log message type for first 5 messages and every 20th message to see what
   message types we're receiving from Agent SDK
2. Log ResultMessage contents: has_result, assistant_msgs count, tool_calls
   count to understand what was captured
3. Log tools used (if any) to verify tool tracking is working

**Next**: Restart bot and retry failing task. Check logs to see:
- What message types are actually being received (expecting AssistantMessage)
- Whether tool_use blocks are present in content
- Why tool_names list is empty despite work being done

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

llm_interface.py

@@ -588,6 +588,11 @@ class LLMInterface:
                     message_count += 1
                     self._last_message_count = message_count
 
+                    # DEBUG: Log message type to understand what we're receiving
+                    msg_type = type(message).__name__
+                    if message_count <= 5 or message_count % 20 == 0:
+                        logger.debug(f"[LLM] Message #{message_count}: {msg_type}")
+
                     if isinstance(message, AssistantMessage) and hasattr(message, 'content'):
                         if isinstance(message.content, str):
                             assistant_messages.append(message.content)
@@ -601,6 +606,11 @@ class LLMInterface:
                                         self._last_tool_names = tool_names.copy()
 
                     if isinstance(message, ResultMessage):
+                        # DEBUG: Log what we captured during message processing
+                        logger.debug(f"[LLM] ResultMessage: has_result={bool(message.result)}, assistant_msgs={len(assistant_messages)}, tool_calls={len(tool_names)}")
+                        if tool_names:
+                            logger.debug(f"[LLM] Tools used: {', '.join(list(dict.fromkeys(tool_names))[:10])}")
+                        
                         result_text = message.result or "\n".join(assistant_messages)
 
                         if not result_text and tool_names: