Repeatability
High
Bash script refactoring follows well-established patterns: retry loops, exit code checks, AWS CLI calls for rollback, and structured logging are all standard, repeatable constructs. The task structure is the same regardless of which specific ECS cluster is targeted.
Ambiguity Tolerance
High
The requirements are unusually specific: retry on network timeout, detailed logging, health-check validation, and a 2-minute rollback window. These are concrete, testable acceptance criteria that leave little room for interpretation.
Data & Tool Availability
Medium
The agent needs the existing bash script and ideally the ECS cluster/service names and task definition naming conventions. It cannot execute against a live AWS environment to validate behavior, so correctness depends on static code review and the agent's knowledge of AWS CLI semantics.
Error Cost
Medium
A buggy rollback script could leave a service in a degraded state or silently fail to revert — but the original problem already has these risks. The refactored script should be reviewed and tested in a staging environment before production use, which limits blast radius.
Human Judgment Required
Low
No taste, ethics, or relationship context is needed. The engineering decisions — retry backoff strategy, how to detect running state, how to fetch the previous task definition ARN — are well-documented AWS patterns with clear best practices.