alerting-dashboard

Alerting Dashboard Skill

Use this skill when configuring or reviewing MCP health dashboards and alert policy.

Workflow

1. Identify the MCP servers that should be monitored and why.
2. Confirm each target name, transport, URL or local executable, and tags before registration.
3. Configure alert thresholds only after explicit approval.
4. Run check_all and review get_dashboard for current health and alert findings.
5. Use get_report for a shareable evidence summary.
6. Document owners, thresholds, expected latency, and known monitoring gaps.

Safety

• Prefer Streamable HTTP for remote targets.
• Treat SSE as legacy compatibility.
• Enable stdio only for trusted local executables.
• Do not change registrations or thresholds without explicit approval.