Add more samples

python/samples/concepts/agents/bedrock_agent/README.md

@@ -10,19 +10,25 @@
 | Sample | Description |
 |--------|-------------|
 | [bedrock_agent_simple_chat.py](bedrock_agent_simple_chat.py) | Demonstrates basic usage of the Bedrock agent. |
-| [bedrock_agent_simple_chat_with_kernel_function.py](bedrock_agent_simple_chat_with_kernel_function.py) | Shows how to use the Bedrock agent with a kernel function. |
-| [bedrock_agent_simple_chat_with_code_interpreter.py](bedrock_agent_simple_chat_with_code_interpreter.py) | Example of using the Bedrock agent with a code interpreter. |
+| [bedrock_agent_simple_chat_streaming.py](bedrock_agent_simple_chat_streaming.py) | Demonstrates basic usage of the Bedrock agent with streaming. |
+| [bedrock_agent_with_kernel_function.py](bedrock_agent_with_kernel_function.py) | Shows how to use the Bedrock agent with a kernel function. |
+| [bedrock_agent_with_kernel_function_streaming.py](bedrock_agent_with_kernel_function_streaming.py) | Shows how to use the Bedrock agent with a kernel function with streaming. |
+| [bedrock_agent_with_code_interpreter.py](bedrock_agent_with_code_interpreter.py) | Example of using the Bedrock agent with a code interpreter. |
+| [bedrock_agent_with_code_interpreter_streaming.py](bedrock_agent_with_code_interpreter_streaming.py) | Example of using the Bedrock agent with a code interpreter and streaming. |
 | [bedrock_mixed_chat_agents.py](bedrock_mixed_chat_agents.py) | Example of using multiple chat agents in a single script. |
+| [bedrock_mixed_chat_agents_streaming.py](bedrock_mixed_chat_agents_streaming.py) | Example of using multiple chat agents in a single script with streaming. |
+| [bedrock_agent_update_agent.py](bedrock_agent_update_agent.py) | Example of updating an agent. |
+| [bedrock_agent_use_existing.py](bedrock_agent_use_existing.py) | Example of using an existing agent. |
 
 ## Before running the samples
 
 You need to set up some environment variables to run the samples. Please refer to the [.env.example](.env.example) file for the required environment variables.
 
-### `AGENT_ROLE_AMAZON_RESOURCE_NAME`
+### `BEDROCK_AGENT_AGENT_RESOURCE_ROLE_ARN`
 
 On your AWS console, go to the IAM service and go to **Roles**. Find the role you want to use and click on it. You will find the ARN in the summary section.
 
-### `FOUNDATION_MODEL`
+### `BEDROCK_AGENT_FOUNDATION_MODEL`
 
 You need to make sure you have permission to access the foundation model. You can find the model ID in the [AWS documentation](https://docs.aws.amazon.com/bedrock/latest/userguide/models-supported.html). To see the models you have access to, find the policy attached to your role you should see a list of models you have access to under the `Resource` section.
 

python/samples/concepts/agents/bedrock_agent/bedrock_agent_update_agent.py

@@ -0,0 +1,56 @@
+# Copyright (c) Microsoft. All rights reserved.
+
+import asyncio
+import uuid
+
+from semantic_kernel.agents.bedrock.bedrock_agent import BedrockAgent
+
+# This sample shows how to update an existing Bedrock agent.
+# This sample uses the following main component(s):
+# - a Bedrock agent
+# You will learn how to connect to a Bedrock agent and update its properties.
+
+
+# Make sure to replace AGENT_NAME and AGENT_ID with the correct values
+AGENT_NAME = "semantic-kernel-bedrock-agent"
+INSTRUCTION = "You are a friendly assistant but you don't know anything about AI."
+NEW_INSTRUCTION = "You are a friendly assistant and you know a lot about AI."
+
+
+async def main():
+    bedrock_agent = await BedrockAgent.create(AGENT_NAME, instructions=INSTRUCTION)
+
+    async def ask_about_ai():
+        new_session_id = str(uuid.uuid4())
+        async for response in bedrock_agent.invoke(
+            session_id=new_session_id,
+            input_text="What is AI in one sentence?",
+        ):
+            print(response)
+
+    try:
+        print("Before updating the agent:")
+        await ask_about_ai()
+
+        await bedrock_agent.update_agent(instruction=NEW_INSTRUCTION)
+
+        print("After updating the agent:")
+        await ask_about_ai()
+    finally:
+        # Delete the agent
+        await bedrock_agent.delete_agent()
+
+    # Sample output (using anthropic.claude-3-haiku-20240307-v1:0):
+    # Before updating the agent:
+    # I apologize, but I do not have any information about AI or the ability to define it.
+    # As I mentioned, I am a friendly assistant without any knowledge about AI. I cannot
+    # provide a definition for AI in one sentence. If you have a different question I can
+    # try to assist with, please let me know.
+    # After updating the agent:
+    # AI is the field of computer science that aims to create systems and machines that can
+    # perform tasks that typically require human intelligence, such as learning, problem-solving,
+    # perception, and decision-making.
+
+
+if __name__ == "__main__":
+    asyncio.run(main())

python/samples/concepts/agents/bedrock_agent/bedrock_agent_use_existing.py

@@ -0,0 +1,54 @@
+# Copyright (c) Microsoft. All rights reserved.
+
+import asyncio
+import uuid
+
+from semantic_kernel.agents.bedrock.bedrock_agent import BedrockAgent
+
+# This sample shows how to interact with a Bedrock agent in the simplest way.
+# This sample uses the following main component(s):
+# - a Bedrock agent that has already been created
+# You will learn how to connect to an existing Bedrock agent and talk to it.
+
+
+# Make sure to replace AGENT_NAME and AGENT_ID with the correct values
+AGENT_NAME = "semantic-kernel-bedrock-agent"
+AGENT_ID = "..."
+
+
+async def main():
+    bedrock_agent = await BedrockAgent.retrieve(AGENT_ID, AGENT_NAME)
+
+    # Use a uiud as the session id
+    new_session_id = str(uuid.uuid4())
+
+    try:
+        while True:
+            user_input = input("User:> ")
+            if user_input == "exit":
+                print("\n\nExiting chat...")
+                break
+
+            # Invoke the agent
+            # The chat history is maintained in the session
+            async for response in bedrock_agent.invoke(
+                session_id=new_session_id,
+                input_text=user_input,
+            ):
+                print(f"Bedrock agent: {response}")
+    except KeyboardInterrupt:
+        print("\n\nExiting chat...")
+        return False
+    except EOFError:
+        print("\n\nExiting chat...")
+        return False
+
+    # Sample output (using anthropic.claude-3-haiku-20240307-v1:0):
+    # User:> Hi, my name is John.
+    # Bedrock agent: Hello John. How can I help you?
+    # User:> What is my name?
+    # Bedrock agent: Your name is John.
+
+
+if __name__ == "__main__":
+    asyncio.run(main())

python/semantic_kernel/agents/bedrock/bedrock_agent.py

@@ -92,7 +92,6 @@ def __init__(
             raise AgentInitializationException("Failed to initialize the Amazon Bedrock Agent settings.") from e
 
         bedrock_agent_model = BedrockAgentModel(
-            agent_arn=bedrock_agent_settings.agent_resource_role_arn,
             agent_id=id,
             agent_name=name,
             foundation_model=bedrock_agent_settings.foundation_model,
@@ -101,14 +100,13 @@ def __init__(
         prompt_template: PromptTemplateBase | None = None
         if instructions and prompt_template_config:
             raise AgentInitializationException("Cannot set both instructions and prompt_template_config.")
-        if not instructions and not prompt_template_config:
-            raise AgentInitializationException("Either instructions or prompt_template_config must be set.")
         if prompt_template_config:
             prompt_template = TEMPLATE_FORMAT_MAP[prompt_template_config.template_format](
                 prompt_template_config=prompt_template_config
             )
 
         args: dict[str, Any] = {
+            "agent_resource_role_arn": bedrock_agent_settings.agent_resource_role_arn,
             "name": name,
             "agent_model": bedrock_agent_model,
         }
@@ -264,7 +262,7 @@ async def create_agent(
             raise AgentInitializationException("Foundation model is required to create an agent.")
 
         await self._create_agent(
-            self.instructions or await self.format_instructions(self.kernel, self.arguments),
+            self.instructions or await self.format_instructions(self.kernel, self.arguments) or "",
             **kwargs,
         )
 

python/semantic_kernel/agents/bedrock/bedrock_agent_base.py

@@ -41,9 +41,12 @@ class BedrockAgentBase(KernelBaseModel):
     function_choice_behavior: FunctionChoiceBehavior = Field(default=FunctionChoiceBehavior.Auto())
     # Agent Model: stores the agent information
     agent_model: BedrockAgentModel
+    # Agent ARN: The Amazon Resource Name (ARN) of the agent.
+    agent_resource_role_arn: str
 
     def __init__(
         self,
+        agent_resource_role_arn: str,
         agent_model: BedrockAgentModel,
         *,
         function_choice_behavior: FunctionChoiceBehavior | None = None,
@@ -54,13 +57,15 @@ def __init__(
         """Initialize the Bedrock Agent Base.
 
         Args:
+            agent_resource_role_arn: The Amazon Resource Name (ARN) of the agent.
             agent_model: The Bedrock Agent Model.
             function_choice_behavior: The function choice behavior.
             bedrock_runtime_client: The Bedrock Runtime Client.
             bedrock_client: The Bedrock Client.
             kwargs: Additional keyword arguments.
         """
         args = {
+            "agent_resource_role_arn": agent_resource_role_arn,
             "agent_model": agent_model,
             "bedrock_runtime_client": bedrock_runtime_client or boto3.client("bedrock-agent-runtime"),
             "bedrock_client": bedrock_client or boto3.client("bedrock-agent"),
@@ -111,7 +116,7 @@ async def _create_agent(
                     self.bedrock_client.create_agent,
                     agentName=self.agent_model.agent_name,
                     foundationModel=self.agent_model.foundation_model,
-                    agentResourceRoleArn=self.agent_model.agent_arn,
+                    agentResourceRoleArn=self.agent_resource_role_arn,
                     instruction=instruction,
                     **kwargs,
                 ),
@@ -159,7 +164,7 @@ async def _create_agent_alias(self, alias_name: str, **kwargs) -> dict[str, Any]
             return await self._run_in_executor(
                 None,
                 partial(
-                    self.bedrock_client.create_client_alias,
+                    self.bedrock_client.create_agent_alias,
                     agentAliasName=alias_name,
                     agentId=self.agent_model.agent_id,
                     **kwargs,
@@ -186,17 +191,18 @@ async def _update_agent(
             raise ValueError("Agent has not been created. Please create the agent before updating it.")
 
         try:
-            self.agent_model = await self._run_in_executor(
+            response = await self._run_in_executor(
                 None,
                 partial(
-                    self.bedrock_client.update_client,
+                    self.bedrock_client.update_agent,
                     agentId=self.agent_model.agent_id,
-                    agentResourceRoleArn=self.agent_model.agent_arn,
+                    agentResourceRoleArn=self.agent_resource_role_arn,
                     agentName=agent_name,
                     foundationModel=foundation_model,
                     **kwargs,
                 ),
             )
+            self.agent_model = response["agent"]
 
             await self._wait_for_agent_status(BedrockAgentStatus.PREPARED)
         except ClientError as e:
@@ -348,7 +354,7 @@ async def associate_agent_knowledge_base(self, knowledge_base_id: str, **kwargs)
             return await self._run_in_executor(
                 None,
                 partial(
-                    self.bedrock_client.associate_client_knowledge_base,
+                    self.bedrock_client.associate_agent_knowledge_base,
                     agentId=self.agent_model.agent_id,
                     agentVersion=self.agent_model.agent_version,
                     knowledgeBaseId=knowledge_base_id,
@@ -372,7 +378,7 @@ async def disassociate_agent_knowledge_base(self, knowledge_base_id: str, **kwar
             await self._run_in_executor(
                 None,
                 partial(
-                    self.bedrock_client.disassociate_client_knowledge_base,
+                    self.bedrock_client.disassociate_agent_knowledge_base,
                     agentId=self.agent_model.agent_id,
                     agentVersion=self.agent_model.agent_version,
                     knowledgeBaseId=knowledge_base_id,
@@ -394,7 +400,7 @@ async def list_associated_agent_knowledge_bases(self, **kwargs) -> dict[str, Any
             return await self._run_in_executor(
                 None,
                 partial(
-                    self.bedrock_client.list_client_knowledge_bases,
+                    self.bedrock_client.list_agent_knowledge_bases,
                     agentId=self.agent_model.agent_id,
                     agentVersion=self.agent_model.agent_version,
                     **kwargs,

python/semantic_kernel/agents/bedrock/models/bedrock_agent_model.py

@@ -17,7 +17,6 @@ class BedrockAgentModel(KernelBaseModel):
     # This model_config will merge with the KernelBaseModel.model_config
     model_config = ConfigDict(extra="allow")
 
-    agent_arn: str = Field(..., alias="agentArn", description="The Amazon Resource Name (ARN) of the agent.")
     agent_id: str | None = Field(None, alias="agentId", description="The unique identifier of the agent.")
     agent_name: str | None = Field(None, alias="agentName", description="The name of the agent.")
     agent_version: str | None = Field(None, alias="agentVersion", description="The version of the agent.")