feat: add collect traces command

[CasLubbers]

📌 Summary

Updated kubectl apply with --server-side to prevent errors and warnings when applying essential manifets.

---

Adds a new collect traces command that generates a comprehensive report of unhealthy Kubernetes resources during APL installation failures. The report is automatically stored in a ConfigMap (apl-operator/apl-traces-report) for post-mortem analysis.

• Monitors 8 resource types: Pods, Deployments, StatefulSets, Nodes, Services, PVCs, PVs, and ArgoCD Applications
• Automatically runs when install or apply commands fail
• Can be invoked manually: binzx/otomi collect traces
• Stores structured JSON report with timestamp and summary statistic

🔍 Reviewer Notes

• Error handling: ArgoCD checks gracefully skip if CRD doesn't exist (404/403 errors)
• Non-blocking: Troubleshooting errors are caught and logged but don't prevent retries
• ConfigMap storage: Uses create-or-update pattern from existing operator patterns
• Minimal output: Console shows only count; full report stored in ConfigMap
• Integration points: Hooked into retry catch blocks in install.ts:149-154 and apply.ts:90-95

🧹 Checklist

• Code is readable, maintainable, and robust.
• Unit tests added/updated