Preprocessing of the initial question

app.py

@@ -3,6 +3,8 @@ import gradio as gr
 import requests
 import pandas as pd
 
+from core.agent import GaiaAgent
+
 # (Keep Constants as is)
 # --- Constants ---
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
@@ -10,13 +12,14 @@ DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 # --- Basic Agent Definition ---
 # ----- THIS IS WERE YOU CAN BUILD WHAT YOU WANT ------
 class BasicAgent:
+    agent: GaiaAgent
     def __init__(self):
-        print("BasicAgent initialized.")
+        self.agent = GaiaAgent()
     def __call__(self, question: str) -> str:
         print(f"Agent received question (first 50 chars): {question[:50]}...")
-        fixed_answer = "This is a default answer."
-        print(f"Agent returning fixed answer: {fixed_answer}")
-        return fixed_answer
+        answer = self.agent.__call__(question)
+        print(f"Agent returning fixed answer: {answer}")
+        return answer
 
 def run_and_submit_all( profile: gr.OAuthProfile | None):
     """

config/prompts.yaml

@@ -85,4 +85,14 @@ prompts:
     type: memory_optimization
     variables: ["summary"]
     version: 1.0
-    description: "Prompt for summarization and memory optimization"
+    description: "Prompt for summarization and memory optimization"
+  question_evaluation:
+    content: |
+      You are a general AI assistant. You are given with a question. Evaluate this question and answer with a YES if the question involves the analysis of
+      an attached file or video link, otherwise with a NO.
+      This is the question to be evaluated
+      {{question}}
+    type: question_refinement
+    variables: ["question"]
+    version: 1.0
+    description: "Prompt for evaluating a question"

core/agent.py

@@ -1,9 +1,9 @@
 from langchain_core.messages import HumanMessage
 from langgraph.graph.state import CompiledStateGraph
 
-
+from core.edges import file_condition
 from core.state import State
-from nodes.nodes import assistant, optimize_memory, response_processing
+from nodes.nodes import assistant, optimize_memory, response_processing, orchestrator
 from tools.tavily_tools import llm_tools
 
 from langgraph.graph import START, StateGraph, END
@@ -19,13 +19,20 @@ class GaiaAgent:
         builder = StateGraph(State)
 
         # Define nodes: these do the work
+        builder.add_node("orchestrator", orchestrator)
         builder.add_node("assistant", assistant)
         builder.add_node("tools", ToolNode(llm_tools))
         builder.add_node("optimize_memory", optimize_memory)
         builder.add_node("response_processing", response_processing)
 
         # Define edges: these determine how the control flow moves
-        builder.add_edge(START, "assistant")
+        builder.add_edge(START, "orchestrator")
+        builder.add_conditional_edges("orchestrator",
+            # If the question involves a file processing -> file_condition routes to the END state
+            # If the question does not involve a file processing -> tools_condition routes to
+            # assistant
+            file_condition)
+
         builder.add_conditional_edges(
             "assistant",
             # If the latest message (result) from assistant is a tool call -> tools_condition routes to tools If the
@@ -33,6 +40,7 @@ class GaiaAgent:
             # response_processing
             tools_condition, {"tools": "tools", "__end__": "response_processing"}
         )
+
         builder.add_edge("tools", "optimize_memory")
         builder.add_edge("optimize_memory", "assistant")
         builder.add_edge("response_processing", END)
@@ -41,10 +49,11 @@ class GaiaAgent:
     def __call__(self, question: str) -> str:
         messages = [HumanMessage(content=question)]
         messages = self.react_graph.invoke({"messages": messages})
-        for m in messages['messages']:
-            m.pretty_print()
+#        for m in messages['messages']:
+#            m.pretty_print()
 
-        return m
+        answer = messages['messages'][-1].content
+        return answer
 
     def __streamed_call__(self, question: str) -> str:
         # Test the web agent

core/state.py

@@ -4,3 +4,4 @@ from langgraph.graph import MessagesState
 class State(MessagesState):
     summary: str
     question: str
+    attachment: str

nodes/nodes.py

@@ -13,6 +13,23 @@ response_processing_model = ChatOpenAI(model="gpt-4.1-mini")
 model = model.bind_tools(llm_tools, parallel_tool_calls=False)
 
 
+# Node
+def orchestrator(state: State):
+    # Get original question if it exists
+    question = state.get("question", "")
+    if not question:
+        question = state["messages"][0].content
+
+    message = prompt_mgmt.render_template("question_evaluation", {"question": question})
+
+    # Add prompt to our history
+    messages = [HumanMessage(content=message)]
+    response = response_processing_model.invoke(messages)
+    if response.content == "YES":
+        return {"question": question, "attachment": "true", "messages":[response]}
+    return {"question": question}
+
+
 # Node
 def assistant(state: State):
     # set up the question
@@ -22,7 +39,7 @@ def assistant(state: State):
     # Get original question if it exists
     question = state.get("question", "")
     if not question:
-        question = state["messages"][0]
+        question = state["messages"][0].content
 
     sys_msg = SystemMessage(content=prompt_mgmt.render_template("base_system", {"summary": summary}))
     try:
@@ -39,9 +56,7 @@ def assistant(state: State):
 def response_processing(state: State):
     question = state.get("question", "")
     answer = state["messages"][-1]
-    print("Question:", question.content)
-    print("Answer:", answer.content)
-    gaia_messages = [HumanMessage(content=question.content), AIMessage(content=answer.content)]
+    gaia_messages = [HumanMessage(content=question), AIMessage(content=answer.content)]
     gaia_sys_msg = SystemMessage(content=prompt_mgmt.render_template("final_answer_processor", {}))
     response = response_processing_model.invoke([gaia_sys_msg] + gaia_messages)
 
@@ -56,7 +71,7 @@ def optimize_memory(state: State):
     if summary:
 
         # A summary already exists
-        summary_message = prompt_mgmt.render_template("summarization", {"summary":summary})
+        summary_message = prompt_mgmt.render_template("summarization", {"summary": summary})
 
     else:
         summary_message = "Create a summary of the conversation above:"

utils/prompt_manager.py

@@ -11,6 +11,7 @@ class PromptType(Enum):
     BASE_SYSTEM = "base_system"
     ANSWER_REFINEMENT = "answer_refinement"
     MEMORY_OPTIMIZATION = "memory_optimization"
+    QUESTION_REFINEMENT = "question_refinement"
 
 
 @dataclass