update app.py

app.py

@@ -5,6 +5,10 @@ import inspect
 import pandas as pd
 from langchain_core.messages import HumanMessage
 from agent import build_graph 
+from huggingface_hub import hf_hub_download
+import logging
+
+logger = logging.getLogger(__name__)
 
 # (Keep Constants as is)
 # --- Constants ---
@@ -25,6 +29,29 @@ class BasicAgent:
         print(f"Agent returning answer: {answer}")
         return answer
 
+def file_extract(local_file_path, task_id):
+    if not local_file_path:
+        return None
+        
+    token = os.getenv("HUGGINGFACEHUB_API_TOKEN") or os.getenv("HF_TOKEN")
+    
+    # GAIA files are usually placed in date-based subdirectories
+    prefixes = ["2023/validation/", "2023/test/", "2023/train/", ""]
+    
+    for prefix in prefixes:
+        try:
+            resolved_path = hf_hub_download(
+                repo_id="gaia-benchmark/GAIA",
+                filename=f"{prefix}{local_file_path}",
+                repo_type="dataset",
+                token=token
+            )
+            return resolved_path
+        except Exception:
+            continue
+            
+    logger.warning(f"Could not download file '{local_file_path}' for task_id {task_id}. Make sure you accepted GAIA terms on HF and set HF_TOKEN.")
+    return None
 
 def run_and_submit_all( profile: gr.OAuthProfile | None):
     """
@@ -90,7 +117,12 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
             continue
             
         if file_name:
-            question_text += f"\n\n[Attached File: {file_name}]"
+            # Actually download the file to local cache and get absolute path
+            resolved_path = file_extract(file_name, task_id)
+            if resolved_path:
+                question_text += f"\n\n[Attached File Local Path: {resolved_path}]"
+            else:
+                question_text += f"\n\n[Attached File: {file_name} (Download Failed)]"
         try:
             submitted_answer = agent(question_text)
             answers_payload.append({"task_id": task_id, "submitted_answer": submitted_answer})