Add system test for Lambda executor Dead Letter Queue (DLQ) processing

[isaiahiruoha]

This PR adds a system test for AWS Lambda executor's Dead Letter Queue (DLQ) functionality, addressing a gap in test coverage for Lambda service-level failures.

Before, all Lambda executor system tests only covered the "happy path" where tasks execute successfully (whether the test ultimately passes or fails). However, the Lambda executor has unique handling for cases where the Lambda service itself fails to complete invocation (timeouts, memory limits, crashes, etc.). In these scenarios:

1. Normal path: Lambda completes → sends result to results queue → executor processes
2. DLQ path: Lambda fails at service level → AWS sends message to DLQ → executor processes DLQ message → marks task as failed

The test uses executor_config functionality (from #53994) to implement a "poison pill" pattern:

1. Task definition: Uses @task(executor_config={"poison_pill": True}) to flag a task for intentional failure
2. Lambda handler: Checks for poison pill flag and raises RuntimeError before any task execution
3. DLQ trigger: RuntimeError causes Lambda service failure → message sent to DLQ
4. Verification: Polls CloudWatch SQS metrics to confirm DLQ processing occurred

Testing: Tested locally on Lambda executor: Confirms complete DLQ flow works correctly

---

^ Add meaningful description above
Read the Pull Request Guidelines for more information.
In case of fundamental code changes, an Airflow Improvement Proposal (AIP) is needed.
In case of a new dependency, check compliance with the ASF 3rd Party License Policy.
In case of backwards incompatible changes please leave a note in a newsfragment file, named {pr_number}.significant.rst or {issue_number}.significant.rst, in airflow-core/newsfragments.