fix(memory-integ): address capacity cap and update failures in tests

[Hweinstock]

Failing integ tests on main: https://github.com/aws/bedrock-agentcore-sdk-python/actions/runs/30303039785/job/90288106089.

Problem

Failing integ tests on main: https://github.com/aws/bedrock-agentcore-sdk-python/actions/runs/30303039785/job/90288106089 with two root causes:

• add_strategy(wait_for_active=True) only waited for the strategy to reach ACTIVE, but the memory itself could still be in UPDATING state. Consecutive add_strategy calls (as in test_add_strategy_immediately_after_another) would hit ValidationException: Memory is in transitional state UPDATING.
• CI account hit memory limit with 150.

Solution

After the strategy is confirmed ACTIVE, add_strategy now also checks the memory status and waits for it to return to ACTIVE using the existing _wait_for_memory_active utility.

Testing

• All 6 integ tests pass against dev account (previously test 4 failed consistently)

Notes

there appears to be an issue with PRs from forks: #603 so make this from origin branch.

[github-actions]

✅ No Breaking Changes Detected

No public API breaking changes found in this PR.

[agentcore-devx-automation]

Claude Security Review: no high-confidence findings. (run)

[Hweinstock]

FAILED tests_integ/evaluation/test_dataset_client.py::TestDatasetCrud::test_create_dataset_and_wait_inline - botocore.errorfactory.ConflictException: An error occurred (ConflictException) when calling the CreateDataset operation: A dataset with this name already exists in your account. Please choose a different name and try again.

evaluator failure is likely a flaky test, can track as a follow-up.

[agentcore-devx-automation]

Claude Security Review: no high-confidence findings. (run)