The job: Your AI-generated project broke, the error message means nothing to you, and you’re going round in circles.
What AI actually does here
- Translates the error into plain English.
- Identifies the likely cause and proposes the smallest fix.
- Suggests how to test that it’s really fixed.
Tools to look at
- AI code editors and assistants: Cursor, GitHub Copilot
- General assistants (ChatGPT, Claude, Gemini)
Named tools are examples, not endorsements. We have no paid relationships with them at the moment; see our disclosure.
What you need first
- Remove passwords, API keys and customer data before pasting anything.
- Use version control so you can roll back a bad fix.
How to measure ROI
- Time stuck per bug, before and after.
- Bugs that come back (a sign of patching symptoms rather than causes).
Run the numbers in the ROI calculator →
What the evidence says
GitHub’s own research reports faster task completion with Copilot (vendor research). Debugging gains vary widely.
Ready-to-use prompt
Replace the [BRACKETS]. Only paste data you’re allowed to share with the tool you’re using.
View promptI'm not an experienced developer. Here is an error and the relevant code.
I'm not an experienced developer. Here is an error and the relevant code.
ERROR:
"""
[paste full error]
"""
CODE (file name and contents):
"""
[paste]
"""
What I was trying to do: [one sentence]
What changed just before it broke: [if known]
1. Explain the error in plain English.
2. Give the most likely cause.
3. Show the smallest change that fixes it.
4. Tell me how to check it worked.
Don't rewrite unrelated code.Sources
- Cursor (product page) Vendor page
- GitHub, "Research: quantifying GitHub Copilot's impact on developer productivity and happiness" Vendor research
Last updated . Spotted something out of date? Email support@aiforthisjob.com. How we research and check playbooks.