Practice debugging failed CI runs before they block a release.
Start from red pipeline logs, inspect the workflow and config, identify the failing step, and choose the safest fix direction. Each run is a curated red pipeline — you read the evidence and build a diagnosis. No terminal, no live pipeline.
Failed pipeline runs
Tests pass on your laptop, but CI dies during the build. The pipeline is running an older Node than the project declares it needs.
The suite is green locally but red in CI. A required environment variable never reaches the CI job, and the config loader fails loudly when it is absent.
The container build fails on a COPY step. The Dockerfile still copies an output folder the project no longer produces after a build-config change.
How a diagnosis is scored
Each run is graded objectively out of 100: mark the right suspect artifact (25), pick the failing step (25), name the root-cause category (20), choose the safe fix direction (20), and reference a real signal in your note (10). After you submit, the expected diagnosis and a Buglyst reference diagnosis are revealed so you can compare your read of the failure.