Kept asking a person to run a check the AI could have run itself

Computer maintenance

Situation: To check what a setting contained, the AI asked the user to carry out a step that involved typing a long input.
What the AI did: It had the user enter a long command, but a line wrapped partway and broke the input, four failures in a row. When the AI ran it itself, it quickly found the setting's value was empty.
How to ask next time (the operator's example): "Before asking me, try running it on your side first. If you really can't, tell me why when you ask. If the same approach fails twice, switch to a different one."
One question to ask: Can the AI run this itself?
Check it yourself: When an AI asks you to type or paste a long command, first ask: "Can you run this yourself?" Many coding agents can run shell commands on their own. If it says no, ask why.
AI used (month): Claude Code (2026-05)
Times this mistake was recorded: once
Source: The operator's work records (the details were changed to a different task)