The system nobody wants to touch.
The scenario:
Your client has a server running an application that nobody understands anymore. It’s been working for nine years. Nobody touches it. The engineer who built it left in 2021.
It’s also running an unsupported OS, has no backup tested in two years, and one disk is showing SMART warnings. The reason nobody touches it is the reason it’s the biggest risk in the environment. Technical debt isn’t about old code. It’s about unowned systems that hold up the business until they don’t.
The prompt:
You’re building a quarterly technical debt review for your client environments.
Works in any AI tool.
Context: [your CMDB or asset inventory, patch compliance reports, incident reports from the past 6 months]
Build:
A scoring method factoring age, supportability, business criticality, single point of failure, knowledge concentration
The top 10 systems with the highest debt score per client
A status for each: investing, replacing, retiring, accepting
A written risk acknowledgement requirement for anything marked accepting
A one-page client summary showing the top 3 risks and the decisions they need to make
Two pages total.
