test-failure-investigator

Azure DevOps pipeline test failure investigation applies a hypothesis-driven methodology that analyzes pipeline run URLs, filters flaky retries, queries Kusto telemetry, and generates ranked root cause hypotheses with validation plans.

Retry filtering works by automatically excluding ADO pipeline tests that pass on retry, ensuring investigations focus strictly on blocking failures that require root cause analysis and fix strategies.

Yes, you can use Kusto queries to identify recent changes causing test failures by analyzing telemetry data retrieved during the investigation to correlate anomalies with pipeline errors and rank potential root causes.

A hypothesis-driven root cause analysis approach formulates potential causes for build failures, ranks them by likelihood, and generates validation plans to systematically test and confirm the exact underlying issue.

Yes, this test failure investigation method works with any Azure DevOps pipeline run URL, analyzing specific build details to apply retry filtering and query Kusto telemetry for that run.

Limitations of automated hypothesis generation for debugging include reliance on available Kusto telemetry and pipeline logs, meaning it may not identify root causes hidden in unlogged external dependencies or local environment issues.