Introducing the “retry” step failure strategy for Bitbucket Pipelines

This post introduces the `retry` step failure strategy for Bitbucket Pipelines, allowing automatic retries of failed steps up to a configurable maximum count (1-10). It also extends failure strategy support to the parent level of parallel groups and stages, enabling inherited strategies that can be overridden at the step level. The `retry` strategy is mutually exclusive with `fail-fast` and does not support deployment steps or stages. Visual updates in the Pipelines view now communicate retry attempts.