The PagerDuty MCP server lets your AIML.chat agents look up incidents and on-call schedules, and create or update incidents from a conversation — with a human approving every write. It installs from the MCP marketplace in a few clicks and gives your technical agent live incident context.
What your agents can do with PagerDuty
Once connected, a technical or support agent can check whether there's an active incident affecting a user's reported problem, read its status to give an accurate "we're on it" answer, and tell the user who's on call. When a new incident is warranted, the agent can draft and trigger one with the right service and urgency — but only with approval, so you never page a team by accident. Reads happen automatically; anything that writes runs through AIML.chat's write-confirm flow, so the agent proposes and a person approves before it touches PagerDuty. Grounded in your own content plus your PagerDuty data, the agent gives accurate, cited answers, and the analytics surface the questions that recur during incidents.
Tools this MCP server exposes
The PagerDuty MCP server exposes read tools — list incidents, read an incident, look up on-call and schedules — and write tools — create or trigger an incident, acknowledge or resolve, add a note. You decide which services and tools each agent may use and scope the connection to least privilege. Read-only tools are safe to enable broadly; write tools always require approval, so no one is paged without sign-off.
Connect it in three steps
- 1In AIML.chat, open MCP → Marketplace → PagerDuty and click Install.
- 2Authorize with PagerDuty (API token) and scope access to the right services.
- 3Attach the PagerDuty tools to your technical agent and test: ask it to "check for active incidents" and, if needed, approve a trigger.
Most teams start with read-only tools to answer accurately, then enable incident actions once they trust the approval flow. You start on a real free tier and scale as usage grows — see pricing for the agent and message limits, and the MCP docs for how write-confirm works. When you're ready, start free.
Will it page a team on its own?+
No. Reads are automatic, but triggering, acknowledging, or resolving an incident requires human approval first.
What can the agent do with PagerDuty?+
Look up incidents and on-call, and create, acknowledge, or resolve incidents — with approval on writes.
Is there a free plan?+
Yes — start on the free tier. See pricing for higher agent and message limits.