feat(tokens): Implement accurate token counting for Gemini history

ai_client.py

@@ -1192,9 +1192,27 @@ def get_history_bleed_stats() -> dict:
             "percentage": percentage,
         }
     elif _provider == "gemini":
-        # For Gemini, token estimation is complex and handled by the server.
-        # We don't have a reliable client-side estimate, so we return a
-        # "not implemented" state for now.
+        if _gemini_chat:
+            try:
+                _ensure_gemini_client()
+                history = _get_gemini_history_list(_gemini_chat)
+                if history:
+                    resp = _gemini_client.models.count_tokens(
+                        model=_model,
+                        contents=history
+                    )
+                    current_tokens = resp.total_tokens
+                    limit_tokens = _GEMINI_MAX_INPUT_TOKENS
+                    percentage = (current_tokens / limit_tokens) * 100 if limit_tokens > 0 else 0
+                    return {
+                        "provider": "gemini",
+                        "limit": limit_tokens,
+                        "current": current_tokens,
+                        "percentage": percentage,
+                    }
+            except Exception:
+                pass
+        
         return {
             "provider": "gemini",
             "limit": _GEMINI_MAX_INPUT_TOKENS,