Examples and Quick Reference
Quick reference for natural language prompts, workflow triggers, and command patterns. For detailed step-by-step guides, see Common Workflows.
Example Requests
On-Call Triage
"Summarize what's firing right now."
"Show me all critical alerts in production."
"How are the current alerts grouped?"
"What's the alert landscape for the checkout service?"
Silence Management
"Silence alerts for the checkout service for 2 hours during deployment."
"Preview what would happen if I silence all api-server alerts."
"Extend the current checkout silence by 30 minutes."
"Expire all my active silences — maintenance is done."
"Would this silence pass our compliance policy?"
Routing & Notifications
"Who gets paged when a critical alert fires for the api-server?"
"What Slack channels are configured for alerts?"
"Are any alerts falling through to the default route?"
"Show me the full routing tree."
Integration Testing
"Push a test alert to verify that the slack-sre receiver is working."
"Fire a test alert with labels team=platform and severity=warning."
"Simulate routing for alertname=MCPIntegrationTest to check the pipeline."
Governance & Audit
"Show me all silence changes in the last 24 hours."
"Who created silences yesterday?"
"Export the current Alertmanager config for Git storage."
"Show the MCP audit log."
Command Reference Guide
| User Intent | Core Tools | Example Query |
|---|---|---|
| On-call summary | am_summarize_oncall | "What's firing right now?" |
| Filter alerts | am_list_alerts | "Show critical alerts for checkout" |
| Check alert groups | am_list_alert_groups | "How are alerts grouped?" |
| Preview silence | am_preview_silence | "What would this silence affect?" |
| Create silence | am_create_silence | "Silence checkout for 2 hours" |
| Extend silence | am_update_silence | "Add 30 more minutes" |
| Expire silence | am_expire_silence | "Maintenance is done, expire it" |
| Silence specific alert | am_silence_alert | "Mute this HighCPU alert" |
| Check routing | am_explain_routing | "Who gets paged for this?" |
| Audit default route | am_audit_default_route | "Any alerts falling through?" |
| Test notification | am_push_test_alert | "Verify Slack integration works" |
| Audit changes | am_list_recent_changes | "Who created silences today?" |
| Validate policy | am_validate_silence_policy | "Would this silence be allowed?" |
Tips and Best Practices
Always Preview Before Silencing
Check the blast radius before creating any silence:
"Preview what would happen if I silence service=checkout in production."
Use Service Scope for Silences
The default service scope (alertname + service + env) is the safest balance between too narrow and too broad:
"Silence the HighCPU alert for api-server in production for 1 hour."
Use Guided Prompts for Complex Workflows
The server includes 3 built-in guided prompts:
| Prompt | Use Case |
|---|---|
am-alert-triage-guided | On-call shift triage |
am-maintenance-silence-guided | Planned maintenance silencing |
am-integration-test-guided | Notification pipeline testing |
Audit Regularly
Review silence activity daily for compliance:
"Show me all silence changes in the last 24 hours and who created them."
Test Alerts Are Real
Remember that am_push_test_alert fires real notifications. Use descriptive labels:
"Push a test alert with alertname=MCPIntegrationTest, team=sre, severity=info."
Next Steps
- Common Workflows — Step-by-step workflow guides
- Tools — Available MCP tools reference
- Resources — MCP resources for monitoring
- Configuration — Server configuration