GitHub Actions Logs

Resolve the repository, run, and optional job from the supplied URL. Preserve the requested attempt when a run has been rerun.

Use an available GitHub connector or authenticated gh. With gh, inspect status before choosing logs (replace OWNER/REPO, RUN_ID, and JOB_ID):

gh run view RUN_ID --repo OWNER/REPO --json status,conclusion,headSha,attempt,jobs
gh run view RUN_ID --repo OWNER/REPO --log-failed

For a particular job, use gh run view --repo OWNER/REPO --job JOB_ID --log. Save large logs to a temporary file and inspect the failing step with context. Report the run/attempt, commit, failure cause, and supporting log excerpt; distinguish retrieval failures from build/test failures.

Access And Availability

• If gh is absent, use an available connector or the GitHub REST API. Installing system packages is not a prerequisite for this workflow.
• Logs for an active job may not be available yet. If completed-job logs are available, inspect those; otherwise report the available status and the missing evidence. Do not assume the REST endpoint provides live logs.
• Fetch a supplied signed log URL with an available HTTP tool. Treat its query string as a credential; do not echo it in reports. If it expires, retrieve a fresh download through the run/job API when access permits.
• If access is missing, report what could be inspected and the specific access needed. Never print credentials or silently fall back to another run.
• Log inspection does not authorize rerunning, cancelling, or deleting runs.

CLI options and log retrieval limitations are documented in gh run view.