AI Prompt for Decoding Error Messages

Decode cryptic error messages with this AI prompt: plain-English translation, root-cause explanation, safest fix ranked by risk, and verification steps.

The prompt

Role: You are a developer advocate who translates cryptic error messages into plain English and safe fixes.

Context: I was trying to [WHAT_YOU_WERE_DOING] when this happened. My setup: [ENVIRONMENT: OS, tool versions].

Full error message:
[ERROR_MESSAGE]

Task: 1) Translate the error into plain English: what is the computer complaining about? 2) Explain the root cause — the underlying situation that produces this error, not just the symptom. 3) Give the fix as numbered steps, ordered from safest to most drastic. 4) Warn me about the common WRONG fix people try for this error (e.g. chmod 777, --force) and why it is dangerous. 5) Tell me how to verify the fix worked.

Format: Translation first (one paragraph), then root cause, then fix steps, then the warning box, then verification.

Ask the AI to show you how to reproduce the error on demand in a minimal setup, so you can confirm the fix actually resolves it instead of masking it. Reproducible errors are debuggable errors.

Constraints: Explain every command before telling me to run it. Never suggest disabling security features as the first resort.

Variables to fill in

Replace each [variable] in the prompt with your own details:

[WHAT_YOU_WERE_DOING] [ENVIRONMENT] [ERROR_MESSAGE]

Example input

ERROR=Error: EACCES: permission denied, open '/usr/local/lib/node_modules', CONTEXT=trying to install a global npm package on macOS

How to use this prompt

Paste the FULL error including the first line — the error code carries the most signal. Say what you were doing; the same error means different things in different contexts.

Tips for the best results

Always read the 'wrong fix' warning — it saves you from Stack Overflow's most dangerous answers.
Ask what the error code (EACCES, ENOENT, etc.) literally stands for to build intuition.
Save translations of errors you hit often as team documentation.