What problem does it solve? Diagnosing production incidents in the Adobe Dispatcher Apache HTTP Server layer of AEM as a Cloud Service is slow and error-prone when done by guesswork. This Skill provides an evidence-first workflow that correlates metrics, logs, request traces, and cache state to isolate root causes and propose safe containment and rollback. ## Core Features & Use Cases - Evidence-First Triage Workflow: Quantify impact with monitor_metrics, gather logs with tail_logs, trace requests with trace_request, and inspect cache state before concluding root cause. - Ten Incident Playbooks: Pre-built investigation paths for 5xx spikes, cache miss regressions, sudden 403 blocks, redirect loops, latency regressions, vanity URL outages, and probe/readiness failures. - Cloud Guardrails: Enforces AEMaaCS-specific rules such as reserved probe paths (/systemready, /system/probes/*), validator contracts, and CDN-vs-Dispatcher decision boundaries. - Use Case: When a deploy causes a 403 spike on legitimate URLs, trace the failing request, identify the exact filter rule denying it, and produce an incident report with probable cause, confidence level, containment, and rollback steps. ## Quick Start Ask the assistant to investigate a Dispatcher incident, for example: "We are seeing a spike of 5xx errors after the last deploy, run a full incident investigation on our AEM cloud dispatcher."
