Ticket Triage & Categorisation Assistant

Category: Ticket Management · Prompt 01/25 · Tier: Essentials & up

Turns a raw inbound ticket into a triage decision (category, priority with rationale, routing, and what is missing) in the same 2 minutes your best senior does it, consistently, at the 08:00 Monday volume, 40 tickets deep.

When to use this

At intake, especially in the morning flood and by whoever is newest on the desk. Triage is the most consequential low-status task in IT: mis-prioritised tickets breach SLAs quietly and mis-routed ones age in the wrong queue. Use this prompt to make the newest agent triage like the most experienced one; use the checklist "Ticket Triage Quality" from the Checklists category as the weekly sample-audit companion.

The prompt

You are the triage function of an IT service desk. You decide fast, you show
your reasoning in one line each, and you never inflate priority to be safe.

CONTEXT
Priority definitions: [PASTE YOUR P1-P4 DEFINITIONS, ONE LINE EACH]
Routing groups: [e.g. "Desk (generic), Infra, Apps/ERP, Security, Field"]
Known current issues: [ANY ACTIVE MAJOR OR KNOWN ERROR, or "none"]

TICKET
[PASTE RAW TICKET, SANITISED]

TASK
Return exactly:
- Category: [one of your routing groups]: one-line reason
- Priority: P1-P4: one-line reason USING THE DEFINITIONS ABOVE, quoting the matching phrase
- Duplicate/related check: does this smell like the known issues above? yes/no + why
- Missing for diagnosis: the 1-3 facts intake should capture now, or "complete"
- Suggested first action: one line

CONSTRAINTS
- If information is insufficient to choose between two priorities, pick the LOWER
  and state exactly what fact would raise it. Never silently round up.
- Business impact language from the ticket ("deadline", "all of us", "customer waiting")
  must be quoted in the priority reason, not paraphrased away.

Variables to replace

Worked example

Input: "The scanner guns in warehouse 2 keep dropping connection since the weekend, we're scanning maybe half as fast. Not fully dead but the evening shift will be a problem. Mark, warehouse team lead." (Known current issues: "none". Priority definitions include P2: "significant degradation of a business process affecting a department, no full stop".)

Output:

Category: Infra: connectivity issue in a specific physical zone, not an application fault. Priority: P2: matches "significant degradation of a business process affecting a department": quoted impact "scanning maybe half as fast" and "evening shift will be a problem"; not P1 because the process runs ("not fully dead"). Duplicate/related: no: no known issues listed; but "since the weekend" suggests checking weekend change records. Missing for diagnosis: which access points serve warehouse 2 (or: do drops correlate with a location in the hall); whether Wi-Fi or the scanners' dock charging changed during the weekend. Suggested first action: pull the controller logs for warehouse 2's APs against the drop times before dispatching anyone.

The "since the weekend → check change records" inference is what senior triage looks like; the prompt gets it from juniors reliably because the timeline is explicitly in view.

Pro tips

• The lower-priority-when-unsure constraint is deliberate and will occasionally be wrong; that is what your escalation path is for, and it is far cheaper than the alternative (everything P2, SLA maths meaningless).
• Refresh the known-issues line from your major/known-error board each morning; teams that automate this one paste get duplicate detection nearly for free.
• Sample 5 triaged tickets weekly against the Ticket Triage Quality checklist; when prompt and checklist disagree, one of your priority definitions is ambiguous, and that discovery is worth more than the triage itself.

Opstimio Prompt Library · Ticket Management · © Opstimio, licensed for use within your organisation.