1
首先写一份精确的错误报告
在向AI询问任何问题之前，写下你期望发生的事情、实际发生的事情以及完整的错误信息。
Example prompt
Help me debug this issue. Expected behavior: [WHAT SHOULD HAPPEN]. Actual behavior: [WHAT IS HAPPENING]. Error message: [FULL ERROR/STACK TRACE]. Relevant code: [PASTE CODE]. Environment: [LANGUAGE/FRAMEWORK/VERSION]. What are the most likely causes and how do I diagnose each one?
2
请AI解释错误并追踪原因
粘贴堆栈跟踪，让AI用简单语言解释它，然后追溯哪一行触发了它以及为什么。
Example prompt
Explain this stack trace in plain English. Identify: the root error (not the symptom), which function/line actually caused it, the chain of calls that led there, and what state the program was in when it crashed. Stack trace: [PASTE STACK TRACE]
3
生成最小重现和修复方案
让AI编写能重现错误的最小可能测试用例。重现后，请求带有解释的修复方案。
Example prompt
Based on this bug description, write: 1) A minimal code snippet that reproduces the issue (no dependencies beyond what's needed). 2) The fix with a clear explanation of why the original code was wrong. 3) A unit test that would have caught this bug. Bug: [DESCRIBE THE BUG]. Relevant code: [PASTE CODE]
Ready to try Blink?
Follow this playbook with the actual tool