메뉴
Use when the user says they did a fresh prod deployment and wants to know if everything is okay. Checks pods, deployments, and events on the jotnar-ymir OpenShift cluster.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Backport upstream patches to packages in the RHEL ecosystem — cherry-pick or git-am workflow, build verification, changelog, and merge request creation.
Rebase packages to newer upstream versions in the RHEL ecosystem — version update, spec file modification, patch fixing, build verification, changelog, and merge request creation.
Triage Jira issues for RHEL packages — analyze bugs and CVEs to determine whether to rebase, backport a patch, rebuild, or request clarification, check CVE applicability against package source, consolidate rebuild siblings, and post the result as a Jira comment.
Runs the Errata Workflow for an erratum, advancing it through states (NEW_FILES -> QE -> REL_PREP), handling stage pushes, CAT test timeouts, product listing verification, and flagging for human attention.
Runs the Issue Verification workflow for a JIRA issue, managing the lifecycle from merged MR through errata creation, testing analysis, and status transitions.
Analyze GreenWave gating and MR OSCI results for a RHEL JIRA issue to determine preliminary testing status, then update JIRA fields or flag attention accordingly.
| name | verify-prod-deployment |
| description | Use when the user says they did a fresh prod deployment and wants to know if everything is okay. Checks pods, deployments, and events on the jotnar-ymir OpenShift cluster. |
Check the health of the jotnar-ymir OpenShift cluster after a fresh deployment. Run these commands in order and report findings.
git log --oneline -25
oc get pods -n jotnar-ymir--jotnar-ymir
Expected: All agent/service pods Running, completed jobs Completed, zero Error or CrashLoopBackOff or ImagePullBackOff.
oc get deploy -n jotnar-ymir--jotnar-ymir
Expected: All deployments show READY count matches desired (e.g. 2/2, 1/1).
oc get imagestream -n jotnar-ymir--jotnar-ymir
Note the UPDATED timestamp for each imagestream. In the report, call out any agent or mcp-server imagestream updated within the last 24 hours — these are the ones that just rolled out.
oc get events -n jotnar-ymir--jotnar-ymir --sort-by='.lastTimestamp' | tail -40
Look for Warning severity lines — Normal is noise. Flag anything other than routine Pulling/Pulled/Created/Started/Completed.
Report as a short table or bullet list:
Recreate (not RollingUpdate). During rollout there is a brief gap with 0 pods — normal, not a failure.FailedCreate + quota error, check oc get appliedclusterresourcequotas. Usually resolves itself with Recreate strategy.Completed status is expected and healthy.