cleanup streamlit examples

Author: cornelcroi

examples/streamlit/imports.py examples/python/imports.pyexamples/streamlit/imports.py renamed to examples/python/imports.py

@@ -3,5 +3,5 @@
 # caution: path[0] is reserved for script path (or '' in REPL)
 
 # import your demo here
-sys.path.insert(1, '../movie-production')
-sys.path.insert(1, '../travel-planner')
+sys.path.insert(1, './movie-production')
+sys.path.insert(1, './travel-planner')

examples/python/main-app.py

@@ -0,0 +1,15 @@
+import imports
+import streamlit as st
+
+st.set_page_config(
+    page_title="AWS Multi-Agent Orchestrator Demos",
+    page_icon="👋",
+)
+
+pg = st.navigation(
+    [
+        st.Page("pages/home.py", title="Home", icon="🏠"),
+        st.Page("movie-production/movie-production-demo.py", title="AI Movie Production Demo" ,icon="🎬"),
+        st.Page("travel-planner/travel-planner-demo.py", title="AI Travel Planner Demo" ,icon="✈️"),
+    ])
+pg.run()

…ovie-production/movie-production-demo.py …ovie-production/movie-production-demo.pyexamples/movie-production/movie-production-demo.py renamed to examples/python/movie-production/movie-production-demo.py examples/movie-production/movie-production-demo.py renamed to examples/python/movie-production/movie-production-demo.py

@@ -2,7 +2,9 @@
 import asyncio
 import streamlit as st
 import os
-from  search_web import tool_handler
+import boto3
+from botocore.exceptions import NoRegionError, NoCredentialsError, PartialCredentialsError
+from search_web import tool_handler
 from tool import Tool
 from multi_agent_orchestrator.orchestrator import MultiAgentOrchestrator, OrchestratorConfig
 from multi_agent_orchestrator.agents import (
@@ -14,21 +16,40 @@
 from multi_agent_orchestrator.classifiers import ClassifierResult
 from supervisor_agent import SupervisorAgent, SupervisorAgentOptions
 
+# Function to test AWS connection
+def test_aws_connection():
+    """Test the AWS connection and return a status message."""
+    try:
+        # Attempt to create an S3 client as a test
+        boto3.client('s3').list_buckets()
+        return True
+    except Exception as e:
+        print(f"Incomplete AWS credentials. Please check your AWS configuration.")
+
+    return False
+
 # Set up the Streamlit app
 st.title("AI Movie Production Demo 🎬")
-st.caption("Bring your movie ideas to life with the teams of script writing and casting AI agents")
+st.caption("Bring your movie ideas to life with AI Movie Production by collaborating with AI agents powered by Anthropic's Claude for script writing and casting.")
 
+# Check AWS connection
+if not test_aws_connection():
+    st.error("AWS connection failed. Please check your AWS credentials and region configuration.")
+    st.warning("Visit the AWS documentation for guidance on setting up your credentials and region.")
+    st.stop()
 
+# Define the tools
 search_web_tool = Tool(name='search_web',
-                          description='Search Web for information',
-                          properties={
-                              'query': {
-                                  'type': 'string',
-                                  'description': 'The search query'
-                              }
-                          },
-                          required=['query'])
-
+                       description='Search Web for information',
+                       properties={
+                           'query': {
+                               'type': 'string',
+                               'description': 'The search query'
+                           }
+                       },
+                       required=['query'])
+
+# Define the agents
 script_writer_agent = BedrockLLMAgent(BedrockLLMAgentOptions(
     model_id='us.anthropic.claude-3-sonnet-20240229-v1:0',
     name="ScriptWriterAgent",
@@ -37,10 +58,11 @@
 develop a compelling script outline with character descriptions and key plot points.
 
 Your tasks consist of:
-1. Write a script outline with 3-5 main characters and key plot points
+1. Write a script outline with 3-5 main characters and key plot points.
 2. Outline the three-act structure and suggest 2-3 twists.
-3. Ensure the script aligns with the specified genre and target audience
-"""))
+3. Ensure the script aligns with the specified genre and target audience.
+"""
+))
 
 casting_director_agent = BedrockLLMAgent(BedrockLLMAgentOptions(
     model_id='anthropic.claude-3-haiku-20240307-v1:0',
@@ -51,24 +73,23 @@
 
 Your tasks consist of:
 1. Suggest 1-2 actors for each main role.
-2. Check actors' current status using search_web tool
+2. Check actors' current status using the search_web tool.
 3. Provide a brief explanation for each casting suggestion.
 4. Consider diversity and representation in your casting choices.
-5. Provide a final response with all the actors you suggest for the main roles
+5. Provide a final response with all the actors you suggest for the main roles.
 """,
-
-tool_config={
-    'tool': [search_web_tool.to_bedrock_format()],
-    'toolMaxRecursions': 20,
-    'useToolHandler': tool_handler
+    tool_config={
+        'tool': [search_web_tool.to_bedrock_format()],
+        'toolMaxRecursions': 20,
+        'useToolHandler': tool_handler
     },
     save_chat=False
 ))
 
 movie_producer_supervisor = BedrockLLMAgent(BedrockLLMAgentOptions(
     model_id='us.anthropic.claude-3-5-sonnet-20241022-v2:0',
     name='MovieProducerAgent',
-    description="""
+    description="""\
 Experienced movie producer overseeing script and casting.
 
 Your tasks consist of:
@@ -77,7 +98,7 @@
 3. Summarize the script outline and casting suggestions.
 4. Provide a concise movie concept overview.
 5. Make sure to respond with a markdown format without mentioning it.
-""",
+"""
 ))
 
 supervisor = SupervisorAgent(SupervisorAgentOptions(
@@ -86,25 +107,23 @@
     trace=True
 ))
 
+# Define async function for handling requests
+async def handle_request(_orchestrator: MultiAgentOrchestrator, _user_input: str, _user_id: str, _session_id: str):
+    classifier_result = ClassifierResult(selected_agent=supervisor, confidence=1.0)
 
-
-async def handle_request(_orchestrator: MultiAgentOrchestrator, _user_input:str, _user_id:str, _session_id:str):
-    classifier_result=ClassifierResult(selected_agent=supervisor, confidence=1.0)
-
-    response:AgentResponse = await _orchestrator.agent_process_request(_user_input, _user_id, _session_id, classifier_result)
+    response: AgentResponse = await _orchestrator.agent_process_request(_user_input, _user_id, _session_id, classifier_result)
 
     # Print metadata
     print("\nMetadata:")
     print(f"Selected Agent: {response.metadata.agent_name}")
     if isinstance(response, AgentResponse) and response.streaming is False:
         # Handle regular response
         if isinstance(response.output, str):
-            return (response.output)
+            return response.output
         elif isinstance(response.output, ConversationMessage):
-                return (response.output.content[0].get('text'))
-
+            return response.output.content[0].get('text')
 
-# Initialize the orchestrator with some options
+# Initialize the orchestrator
 orchestrator = MultiAgentOrchestrator(options=OrchestratorConfig(
     LOG_AGENT_CHAT=True,
     LOG_CLASSIFIER_CHAT=True,
@@ -119,12 +138,10 @@ async def handle_request(_orchestrator: MultiAgentOrchestrator, _user_input:str,
 USER_ID = str(uuid.uuid4())
 SESSION_ID = str(uuid.uuid4())
 
-# Input field for the report query
+# Input fields for the movie concept
 movie_idea = st.text_area("Describe your movie idea in a few sentences:")
-genre = st.selectbox("Select the movie genre:",
-                        ["Action", "Comedy", "Drama", "Sci-Fi", "Horror", "Romance", "Thriller"])
-target_audience = st.selectbox("Select the target audience:",
-                                ["General", "Children", "Teenagers", "Adults", "Mature"])
+genre = st.selectbox("Select the movie genre:", ["Action", "Comedy", "Drama", "Sci-Fi", "Horror", "Romance", "Thriller"])
+target_audience = st.selectbox("Select the target audience:", ["General", "Children", "Teenagers", "Adults", "Mature"])
 estimated_runtime = st.slider("Estimated runtime (in minutes):", 30, 180, 120)
 
 # Process the movie concept
@@ -134,6 +151,7 @@ async def handle_request(_orchestrator: MultiAgentOrchestrator, _user_input:str,
             f"Movie idea: {movie_idea}, Genre: {genre}, "
             f"Target audience: {target_audience}, Estimated runtime: {estimated_runtime} minutes"
         )
-        # Get the response from the assistant
-        response = asyncio.run(handle_request(orchestrator, input_text, USER_ID, SESSION_ID))
-        st.write(response)
+        loop = asyncio.new_event_loop()
+        asyncio.set_event_loop(loop)
+        response = loop.run_until_complete(handle_request(orchestrator, input_text, USER_ID, SESSION_ID))
+        st.write(response)

…e-production/movie-production-result.png …e-production/movie-production-result.pngexamples/movie-production/movie-production-result.png renamed to examples/python/movie-production/movie-production-result.png examples/movie-production/movie-production-result.png renamed to examples/python/movie-production/movie-production-result.png

@@ -1,38 +1,21 @@
 ## 🎬 AI Movie Production Agent
-This Streamlit app is an AI-powered movie production assistant that helps bring your movie ideas to life using Claude 3 on Amazon BedrocK. It automates the process of script writing and casting, allowing you to create compelling movie concepts with ease.
-### streamlit app
-Here is a screenshot of the streamlit app. You can describe your movie, select a movie genre, audience and duration and hit `Develop Movie Concept`
+This Streamlit app is an AI-powered movie production assistant that helps bring your movie ideas to life using Claude 3 on Amazon Bedrock. It automates the process of script writing and casting, allowing you to create compelling movie concepts with ease.
 
+### Streamlit App
+Here is a screenshot of the streamlit app. You can describe your movie, select a movie genre, audience and duration and hit `Develop Movie Concept`
 ![image](./movie-production.png)
 
-After a few seconds you should have a your movie ready! 🍿 🎬
-
+After a few seconds you should have your movie ready! 🍿 🎬
 ![image](./movie-production-result.png)
 
 ### Features
 - Generates script outlines based on your movie idea, genre, and target audience
 - Suggests suitable actors for main roles, considering their past performances and current availability
 - Provides a concise movie concept overview
 
-### How to get Started?
-
-1. Clone the GitHub repository
-
-```bash
-git clone https://github.com/awslabs/multi-agent-orchestrator.git
-```
-2. Install the required dependencies:
-
-```bash
-cd examples/movie-production
-pip install -r requirements.txt
-```
-3. Get your AWS Credentials
+### How to Get Started?
 
-4. Run the Streamlit App
-```bash
-streamlit run movie-production-demo.py
-```
+Check out the [demos README](../README.md) for installation and setup instructions.
 
 ### How it Works?