Update agent.py

agent.py

@@ -10,7 +10,8 @@ from PIL import Image, ImageOps
 import requests
 from io import BytesIO
 from markdownify import markdownify
-import whisper
+import
+ whisper
 import time
 import shutil
 import traceback
@@ -156,7 +157,7 @@ class ExcelReaderTool(Tool):
             "nullable": True
         }
     }
-    output_type = "string"
+    output_type = "pandas.DataFrame"
     
     def forward(self, excel_path: str, sheet_name: str = None) -> str:
         try:
@@ -256,6 +257,8 @@ class MagAgent:
             - Reason step-by-step. Think through the solution logically and plan your actions carefully before answering.
             - Validate information. Always verify facts when possible instead of guessing.
             - When processing external data (e.g., YouTube transcripts, web searches), expect potential issues like missing punctuation, inconsistent formatting, or conversational text.
+            - When asked to process Excel files, use the `excel_reader` tool, which returns a pandas DataFrame.
+            - When calculating sales, make sure you multiply volume on price per each product or category.
             - When asked to transcript YouTube video, try searching it in www.youtubetotranscript.com.
             - If the input is ambiguous, prioritize extracting key information relevant to the question.
             - Use code if needed. For calculations, parsing, or transformations, generate Python code and execute it. Be cautious, as some questions contain time-consuming tasks, so analyze the question and choose the most efficient solution.