Start from the Service Management project template in Manager: requests, incidents, problems and changes get their own flow, and your agent runs the routine.
Each has its own type, statuses and fields, so a password reset never waits behind a migration.
Every incident is graded on one grid, so the worst damage gets fixed first.
Every change has a risk, a date and a backout plan, and normal changes pass the CAB first.
IT service management treats IT as services people rely on and sorts work by kind: requests are fulfilled, breakdowns restored, recurring faults removed at the root, changes controlled.
Problem management finds the cause behind repeated incidents and records known errors with workarounds. Changes are approved by risk: standard ones in advance, normal ones by a change advisory board (CAB), emergencies fast. The template follows ITIL practices; ITIL® is a registered trademark of PeopleCert, and the template is independent of it.
Where each part lives in a Service Management project.
| ITSM | In Manager |
|---|---|
| Service catalog | A wiki page of services and owners, from the service-catalog skill. |
| Impact × urgency | Impact and Urgency fields; the grid in service-catalog turns them into priority. |
| Known error database | Known Error status, field and label on a Problem, plus a wiki page of workarounds. |
| Change types | The Change Type field: Standard skips the CAB, Normal needs it, Emergency is expedited. |
| CAB | The cab-review skill: decisions as comments, the CAB-Approved label, a Scheduled date. |
| Post-implementation review | The Review status and its skill. |
Connect your AI agent over MCP and create the project from the template. Types, filters and skills come with it.
Say to your agentManager: Create a project for our IT from the Service Management template.
List your services, standard requests and owners, and agree on a priority grid. The catalog becomes a wiki page.
Say to your agentManager: Run the service-catalog skill and ask me about our services and who owns them.
Requests get a category and requester. Incidents get impact, urgency and the affected service; a widespread outage is labeled Major Incident.
Say to your agentManager: Open an incident and ask me which service is affected and how badly.
Recurring incidents become a linked problem. Once the cause is clear, it becomes a Known Error with its workaround on the wiki.
Say to your agentManager: Find incidents that keep coming back and open a linked problem for each cause.
Every change gets a type, risk, date and backout plan. Normal ones wait in Assessing for the CAB and get reviewed after going live.
Say to your agentManager: Run the cab-review skill for the changes waiting in Assessing.
A project created from the Service Management template starts with these work item types and statuses. You can adapt all of them to your team later.
A standard, pre-approved request from the service catalog.
An unplanned outage or degradation, resolved to restore service fast.
The root cause behind recurring incidents.
A change to the estate, approved by type and reviewed afterwards.
A unit of work under any of the four.
These skills come with the template in Manager:
service-catalogRecords services, owners and the priority grid. Run it first.
service-requestTakes a standard request from New to Closed.
incident-managementGrades an incident and drives it to restored service.
problem-managementFinds root causes and records known errors.
change-managementTakes a change from assessment to implementation.
cab-reviewChecks changes in Assessing for risk, backout and clashes.
post-implementation-reviewConfirms a change worked, records learnings, closes it.
Incidents here focus on restoring service. For a commander, timeline and postmortem, the incident-management skill points to the Incident Response template.
The service-catalog skill ships an example P1 to P4 grid. Write yours into it; every skill grades by it.
Requester and Owner are project members, and you name the Affected Service in your own words. Due dates bring a reminder and an overdue flag for your response targets.
Saved filters show open incidents, changes and problems, and your agent keeps known errors and the catalog current.
Get early access