| name | check-deployment-status |
| description | Check the health and sync status of all ArgoCD applications across clusters. Identifies out-of-sync, degraded, or unhealthy deployments and provides actionable remediation steps. Use when monitoring deployments, troubleshooting sync failures, or verifying environment health after a release. |
Check Deployment Status
Retrieve and analyze the current state of all ArgoCD-managed applications. Surface any that are `OutOfSync`, `Degraded`, `Missing`, or `Unknown`, and provide clear next steps.
Instructions
- List all ArgoCD applications using the ArgoCD agent's MCP tools. Request application name, project, sync status, health status, and target revision.
- Categorize results into:
- Healthy & Synced - no action needed
- OutOfSync - drift detected between desired and live state
- Degraded / Unhealthy - runtime health issues (crashloops, pending pods, failed probes)
- Missing / Unknown - application or resources cannot be found
- For each unhealthy application, provide:
- Application name and namespace
- Current sync and health status
- Last sync timestamp and result
- Resource-level details (which pods/deployments are failing)
- Suggested remediation (sync, rollback, check resource limits, etc.)
- Summary table at the top with total counts per status category.
Output Format
```markdown
Deployment Status Summary
| Status | Count |
|---|
| Healthy | 42 |
| OutOfSync | 3 |
| Degraded | 1 |
| Missing | 0 |
Issues Requiring Attention
1. my-app (namespace: production)
- Sync Status: OutOfSync
- Health: Degraded
- Last Sync: 2026-02-08T14:23:00Z (Failed)
- Details: Deployment `my-app` has 0/3 ready replicas. Pod `my-app-7d4b8c` in CrashLoopBackOff.
- Recommended Action: Check container logs for startup errors. Verify resource limits and environment variables.
```
Examples
- "Show me the status of all ArgoCD applications"
- "Are there any deployments that are out of sync?"
- "Check deployment health after the last release"
- "Which applications are degraded in the staging cluster?"
Guidelines
- Always paginate results (default page size: 20) to avoid OOM with large application counts
- Group results by project or cluster when there are more than 50 applications
- Highlight time since last successful sync for OutOfSync apps
- If all applications are healthy, confirm with a brief positive summary rather than listing every app
- When degraded apps are found, check for recent deployment events that may correlate
