What problem does it solve? When a Huawei Cloud CCE incident occurs, teams struggle to answer "what changed before the failure". This Skill turns scattered signals—workload rollouts, ConfigMap/Secret metadata, Service, Ingress, NetworkPolicy, RBAC, Node, and cloud-network changes—into an evidence-based causal attribution report with timeline correlation, blast radius, and risk scoring. ## Core Features & Use Cases - Change Candidate Ranking: Builds change candidates from retained ReplicaSet revisions, Kubernetes Events, AOM alarms, metrics, logs, and user-provided change records, then ranks them by temporal order, topology overlap, and response signals. - Blast Radius Mapping: Maps each candidate change to affected Pods, Services, Ingresses/Gateways, Nodes, namespaces, storage objects, and dependency paths. - Structured Evidence Report: Produces a Markdown report with Summary, Change Impact Analysis, Evidence Timeline, Blast Radius, Data Gaps, and Appendix, plus a JSON output schema. - Use Case: After a Deployment rollout, Pods in a namespace start crash-looping. Use this Skill to correlate the rollout history, Events, and AOM alarms, rank the image change as the top risk candidate, and hand off to the workload diagnoser for confirmation. ## Quick Start Analyze whether any recent change in my CCE cluster caused the incident that started at the given fault time, using the specified region, project ID, and cluster ID.
