Mypy Boto3 Bedrock Agent Runtime Stubs

Warnings

• breaking Python 3.8 is no longer supported. Packages built with `mypy-boto3-builder` version 8.12.0 and later (including `mypy-boto3-bedrock-agent-runtime` versions 1.42.3+) explicitly dropped support for Python 3.8.
  Fix: Upgrade your Python environment to version 3.9 or higher.
• breaking TypeDef names might have changed to avoid conflicts. For example, `CreateDistributionRequestTypeDef` might become `CreateDistributionRequestRequestTypeDef` if a method with the same name exists.
  Fix: Review your `TypeDef` imports after upgrading `mypy-boto3-builder` or the specific service stub. Consult the latest generated stub files for the exact type definitions (e.g., in `mypy_boto3_bedrock_agent_runtime.type_defs.pyi`).
• gotcha This library provides only type annotations, not the actual `boto3` implementation. You must install `boto3` separately for your code to run.
  Fix: Ensure `boto3` is installed alongside `mypy-boto3-bedrock-agent-runtime` (e.g., `pip install boto3 mypy-boto3-bedrock-agent-runtime`).
• gotcha For optimal type checking, the `mypy-boto3-*` stub package version should closely match your installed `boto3` version. Significant version discrepancies can lead to incorrect type errors or missed valid types.
  Fix: Keep `mypy-boto3-*` packages updated with your `boto3` version. If `boto3` is updated, ensure `mypy-boto3-*` is also updated to the corresponding version.

Install

• pip install boto3 mypy-boto3-bedrock-agent-runtime Install with boto3

Imports

• BedrockAgentRuntimeClient

  from mypy_boto3_bedrock_agent_runtime.client import BedrockAgentRuntimeClient

• InvokeAgentRequestRequestTypeDef

  from mypy_boto3_bedrock_agent_runtime.type_defs import InvokeAgentRequestRequestTypeDef

  Since `mypy-boto3-builder` 8.9.0, TypeDefs that conflict with method names (e.g., `InvokeAgent`) are renamed with suffixes like `RequestRequestTypeDef` or `RequestExtraTypeDef` to ensure unique names.

Quickstart

This quickstart demonstrates how to use the `mypy-boto3-bedrock-agent-runtime` stubs to type-hint your `boto3` Bedrock Agent Runtime client. It shows how to retrieve a type-hinted client, prepare an `InvokeAgent` request with `TypeDef` annotations, and process the streaming response from an AWS Bedrock Agent. Make sure `boto3` is installed and configure your AWS credentials and region.

import boto3
from typing import TYPE_CHECKING, Dict, Any
import os

# These are for type checking only, not actual runtime imports
if TYPE_CHECKING:
    from mypy_boto3_bedrock_agent_runtime.client import BedrockAgentRuntimeClient
    from mypy_boto3_bedrock_agent_runtime.type_defs import InvokeAgentRequestRequestTypeDef, InvokeAgentResponseTypeDef

def get_bedrock_agent_runtime_client() -> "BedrockAgentRuntimeClient":
    """
    Returns a type-hinted Bedrock Agent Runtime client.
    """
    # In a real application, consider using environment variables or a config file
    # for region and credentials.
    # For this example, we'll assume region is set via AWS_DEFAULT_REGION or similar.
    return boto3.client("bedrock-agent-runtime")

def invoke_agent_example():
    client: BedrockAgentRuntimeClient = get_bedrock_agent_runtime_client()

    # Replace with your actual Agent ID and Alias ID
    # It's recommended to set these as environment variables (e.g., BEDROCK_AGENT_ID)
    agent_id = os.environ.get("BEDROCK_AGENT_ID", "YOUR_AGENT_ID")
    agent_alias_id = os.environ.get("BEDROCK_AGENT_ALIAS_ID", "YOUR_AGENT_ALIAS_ID")
    session_id = "test-session-123" # A unique identifier for the session
    input_text = "What is the weather like in Seattle?"

    if agent_id == "YOUR_AGENT_ID" or agent_alias_id == "YOUR_AGENT_ALIAS_ID":
        print("Please set BEDROCK_AGENT_ID and BEDROCK_AGENT_ALIAS_ID environment variables to run this example.")
        print("You can find these in the AWS Bedrock Agents console.")
        return

    request_params: InvokeAgentRequestRequestTypeDef = {
        "agentId": agent_id,
        "agentAliasId": agent_alias_id,
        "sessionId": session_id,
        "inputText": input_text,
        "enableTrace": True, # Set to False for production to reduce logging
    }

    print(f"Invoking agent with input: '{input_text}'")
    try:
        response: InvokeAgentResponseTypeDef = client.invoke_agent(**request_params)

        # Process the response stream
        response_body = response.get("completion")
        if response_body:
            print("Agent response:")
            for event in response_body:
                if "chunk" in event:
                    chunk = event["chunk"]
                    text = chunk.get("bytes").decode("utf-8")
                    print(f"  {text}", end="")
            print("\n") # Newline after all chunks
        else:
            print("No completion received.")

    except client.exceptions.ResourceNotFoundException as e:
        print(f"Error: Agent or Alias not found. Ensure 'BEDROCK_AGENT_ID' and 'BEDROCK_AGENT_ALIAS_ID' are correct. {e}")
    except client.exceptions.ValidationException as e:
        print(f"Validation Error: {e}")
    except Exception as e:
        print(f"An unexpected error occurred: {e}")

if __name__ == "__main__":
    invoke_agent_example()