Given the test incident set is created
When reviewed
Then it covers at least 4 different platform tags (cdp, databricks, oracle, gcp or aws)
And at least 5 different error classes (OOM, disk_full, network_timeout, auth_failure, pipeline_failure)
Given the test incident set is created
When loaded into the test fixtures
Then all 10 have realistic short_description, long_description, affected_ci, and opened_at fields
Given the test incident set is created
Then corresponding log fixtures exist for at least 8 of the 10 incidents (2 may have no logs to test AC-03 graceful handling)
Linear:
ARI-32Type: ⚙️ Enabler | Sprint: 7 | Points: 3
As a Data Platform OPS engineer, I want a set of 10 test incidents covering diverse platforms and error types so that acceptance criteria can be validated consistently and repeatably.
Acceptance Criteria