capo-iam: typed, async Python SDK for AWS IAM¶

capo-iam is a typed Python SDK for AWS IAM. It has an async client, AsyncIAMClient, for asyncio and trio, and a sync client, IAMClient, with the same methods. Requests and responses are typed with TypedDicts, so your editor and type checker know every field.

It is part of capo, a community-driven alternative to boto3 and aioboto3, built on its own HTTP stack instead of botocore and generated from the official AWS Smithy models. Documentation: https://capo-sdk.dev/services/iam/

Installation¶

pip install capo-iam

Usage¶

from capo_iam import AsyncIAMClient


async def main():
    async with AsyncIAMClient() as iam:
        # Example: call the accept_delegation_request operation
        response = await iam.accept_delegation_request()
        print(response)

Pagination¶

Some operations in this SDK support pagination. If the operation supports pagination it will have an iter_ prefixed method that returns an async iterator.

from capo_iam import AsyncIAMClient


async def main():
    async with AsyncIAMClient() as iam:
        # Example: paginate over get_account_authorization_details
        async for item in iam.iter_get_account_authorization_details():
            print(item)

Waiters¶

Waiters poll an operation until a resource reaches a desired state. If the operation supports waiters it will have a wait_until_ prefixed method.

from capo_iam import AsyncIAMClient


async def main():
    async with AsyncIAMClient() as iam:
        # Example: wait for instance_profile_exists
        await iam.wait_until_instance_profile_exists(max_wait_time=300)

Error Handling¶

The SDK raises exceptions for errors returned by the API. Catch them to handle failures gracefully.

from capo_iam import AsyncIAMClient
from capo_iam.error import ConcurrentModificationException


async def main():
    async with AsyncIAMClient() as iam:
        try:
            await iam.accept_delegation_request()
        except ConcurrentModificationException as e:
            print(f"Error: {e}")
            print(e.data)  # additional error data

Retrying¶

The SDK retries failed operations automatically. Retry behaviour follows the Smithy specification: errors are retried based on their is_retryable and is_throttling_error attributes. Throttling errors use a longer base delay. Network-level failures (connection errors and timeouts) are also retried. Non-retryable errors, such as client errors without the @retryable trait, are raised immediately without further attempts.

The number of attempts defaults to 3 and can be changed at the client level via retry_max_attempts, or per call via config_overrides.

from capo_iam import AsyncIAMClient


async def main():
    async with AsyncIAMClient() as iam:
        # Default: 3 attempts for every operation
        response = await iam.accept_delegation_request()

        # Override per operation
        response = await iam.accept_delegation_request(config_overrides={"retry_max_attempts": 5})

        # Disable retries for this call
        response = await iam.accept_delegation_request(config_overrides={"retry_max_attempts": 1})