Add content fallback for conversation analysis + debug logging

When the LLM returns text instead of a tool call for conversation
analysis, try parsing the content as JSON before giving up. Also
log what the model actually returns on failure for debugging.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

utils/llm_client.py

@@ -478,7 +478,17 @@ class LLMClient:
                                   output_tokens=usage.completion_tokens if usage else None)
                     return self._validate_conversation_result(args)
 
-                logger.warning("No tool call in conversation analysis response.")
+                # Fallback: try parsing the message content as JSON
+                if choice.message.content:
+                    logger.warning("No tool call in conversation analysis — trying content fallback. Content: %s", choice.message.content[:300])
+                    fallback = self._parse_conversation_content_fallback(choice.message.content)
+                    if fallback:
+                        self._log_llm("conversation", elapsed, True, req_json, choice.message.content,
+                                      input_tokens=usage.prompt_tokens if usage else None,
+                                      output_tokens=usage.completion_tokens if usage else None)
+                        return fallback
+
+                logger.warning("No tool call or parseable content in conversation analysis response.")
                 self._log_llm("conversation", elapsed, False, req_json, error="Empty response")
                 return None
 
@@ -512,6 +522,31 @@ class LLMClient:
         result.setdefault("conversation_summary", "")
         return result
 
+    def _parse_conversation_content_fallback(self, text: str) -> dict | None:
+        """Try to parse plain-text content as JSON for conversation analysis."""
+        import re
+
+        # Try direct JSON parse
+        try:
+            result = json.loads(text.strip())
+            if "user_findings" in result:
+                return self._validate_conversation_result(result)
+        except (json.JSONDecodeError, ValueError):
+            pass
+
+        # Try extracting from code block
+        match = re.search(r"```(?:json)?\s*(\{.*?\})\s*```", text, re.DOTALL)
+        if match:
+            try:
+                result = json.loads(match.group(1))
+                if "user_findings" in result:
+                    return self._validate_conversation_result(result)
+            except (json.JSONDecodeError, ValueError):
+                pass
+
+        logger.warning("Could not parse conversation content fallback: %s", text[:200])
+        return None
+
     async def chat(
         self, messages: list[dict[str, str]], system_prompt: str,
         on_first_token=None, recent_bot_replies: list[str] | None = None,