PagerDuty MCP Server: Let AI Manage Incidents and On-Call

# PagerDuty MCP Server: Let AI Manage Incidents and On-Call

PagerDuty is the backbone of incident response for thousands of engineering teams. When your AI agent can check on-call schedules, acknowledge incidents, and query service health, it becomes a force multiplier during outages.

With DataFaucet, you can create a hosted PagerDuty MCP server in 60 seconds. No PagerDuty API token setup, no REST client configuration, no code.

What Your AI Can Do With PagerDuty Access

Once connected, your AI tools gain these capabilities:

• Check who's on-call for any service or escalation policy
• List open incidents with severity, status, and assignee
• Acknowledge or resolve incidents from natural language
• View service health across your entire infrastructure
• Query escalation policies to understand response chains
• Search past incidents for patterns and post-mortems

How to Build It

DataFaucet opens PagerDuty in a cloud browser. It captures every API call behind your clicks, then packages them into MCP tool definitions.

1. Enter app.pagerduty.com in DataFaucet
2. Browse: check on-call, view incidents, look at services, open an escalation policy
3. DataFaucet captures the underlying API traffic
4. Deploy as a hosted MCP server with one click

Connect from any MCP client:

claude mcp add my-pagerduty https://datafaucet.dev/api/mcp/YOUR_SERVER_ID

Example Prompts After Connecting

Once your PagerDuty MCP server is live:

• "Who's on-call for the payments service right now?"
• "List all open P1 incidents and their current status"
• "Acknowledge the database alert, I'm looking into it"
• "What incidents fired for the API gateway this week?"
• "Show me the escalation policy for the platform team"
• "Are there any services in degraded state?"

Why MCP Instead of PagerDuty's Built-In Tools

PagerDuty has Slack integrations and mobile apps. An MCP server gives you something different: AI-native access that works inside your coding environment.

When you're debugging in Cursor and an incident fires, you can acknowledge it, check related services, and pull recent similar incidents without leaving your editor.

Common Workflows

Incident triage: "List all open incidents. Which ones are P1? Who's assigned to each?"

On-call handoff: "Who's on-call for platform after 6pm? Add a note that the Redis issue is being monitored."

Post-mortem research: "Find all incidents for the checkout service in the last 30 days. How many were related to timeouts?"

Status check: "Any open incidents right now? If not, what was the last one resolved and when?"

Combine With Other MCP Servers

The real power comes from combining PagerDuty with your observability stack:

• PagerDuty + Datadog MCP: "There's a P1 for API latency. Pull the Datadog dashboard for that service."
• PagerDuty + Sentry MCP: "Check if the new Sentry errors correlate with the incident timeline."
• PagerDuty + Slack MCP: "Acknowledge the incident and post an update to #incidents."
• PagerDuty + Jira MCP: "Create a follow-up ticket for the root cause and link it to the incident."

Your AI agent becomes the bridge between alerting, observability, and action.

Works Everywhere

Your PagerDuty MCP server connects to Claude Desktop, Cursor, Windsurf, Codex, and any MCP-compatible client. Check incidents from Claude, acknowledge alerts in Cursor, query on-call from Windsurf.

Get Started

Create your PagerDuty MCP server in 60 seconds. Free tier includes 3 servers, no credit card required.

Related: Grafana MCP Server for dashboards, Datadog MCP Server for metrics, Sentry MCP Server for errors, Slack MCP Server for communication, Jira MCP Server for issue tracking.