This prompt describes an agent that helps employees fix common IT problems. It asks what device they are using and what the problem is, from a list such as Wi-Fi, VPN, password resets, printers and software installs, then walks them through three troubleshooting steps for that issue and device. If the problem isn't solved, it collects the employee's name and ID and writes a support ticket summary for the IT team. It was written for Microsoft Copilot Studio but works as an agent description in other builders; creating real tickets needs a separate integration, and the sample four-hour response promise should be replaced with your own.
Create an agent that helps employees troubleshoot common IT issues. Start by asking what type of device they are using, choosing from: Laptop, Desktop, Mobile Phone, or Tablet. Then ask them to select their issue from a list: Cannot connect to WiFi, VPN not working, Password reset needed, Printer not found, Software installation request, or Other. Based on their selection, guide them through three troubleshooting steps specific to that issue and device combination. After each step, ask if the issue is resolved. If resolved, end the conversation with a satisfaction confirmation. If not resolved after all three steps, collect the employee's full name and employee ID, create a support ticket summary, and confirm that the IT team will follow up within 4 business hours.
Turn a support ticket into a known-issue articleThis prompt asks the AI to write a short article for a support knowledge base from a closed ticket conversation. The article has two parts, a short summary of the problem and the fix that was actually provided, with no names of people or companies. It tells the AI to double-check the result for any names before finishing.Customer support
Answer an unhappy customer using checked facts onlyThis prompt has ChatGPT draft a short reply to an unhappy customer. Tags keep the customer's message separate from the facts your team has checked, and the reply may not promise refunds or dates that the facts do not include. If something needed is missing, the assistant lists it before the draft.Customer support
Hand a long ticket history to another team in one summaryThink Hard's prompt condenses a lengthy support conversation for a colleague, such as an engineer or a manager, who has seen none of it. The summary uses fixed headings for the problem, customer impact, a dated timeline, fixes already tried, promises made and the single open question, in under 200 words. Error messages and IDs are copied exactly, and any heading the thread does not cover is marked as missing rather than guessed.Customer support
Ticket triage agent that only routes, never repliesA system prompt for an agent that reads each support ticket and returns one routing decision as JSON, naming the category, priority and queue with a confidence score. You fill in a closed list of categories, priorities, queues and who counts as a VIP. Security, outage, legal and VIP cases jump to the top priority, low-confidence calls go to a human, and unknown fields stay null. Text in a ticket that tries to steer the agent is flagged rather than obeyed.Customer support
