Bits of .NET

🤖 One Prompt Reads the Stack Trace For You

A raw stack trace is technically all the information you need to find a bug’s root cause — but reading twenty frames of framework internals to find the one line that’s actually yours takes real time. A well-scoped AI prompt does that triage instantly.

🎯 The Actual Problem With Stack Traces

A typical production stack trace is 80% framework/library frames
you didn't write and can't fix directly, and 20% your code. Manually
scanning to find where YOUR code first touches the failure, and
what in your code's logic actually caused it, is the slow part -
not reading the exception type itself.

✅ The Prompt

• Paste the FULL stack trace plus the relevant source file(s) — don’t trim the trace yourself, let the model do the filtering:

🤖 Copy This Prompt

Here is a stack trace and the relevant source code. Do the
following, in order:

1. Identify the exact line in MY code (not framework/library code)
   where this failure originates.
2. Explain in plain English, in one paragraph, why this specific
   input or state caused the exception - as if explaining it to
   a teammate who has never seen this code before.
3. Give me the minimal code change that fixes the root cause, not
   just a try/catch that suppresses the symptom.
4. Tell me if this same bug pattern likely exists anywhere else
   in the pasted code.

STACK TRACE:
[PASTE FULL STACK TRACE]

RELEVANT SOURCE:
[PASTE SOURCE FILE(S)]

⚠️ One Thing to Watch For

• Always verify the suggested root cause against the actual line numbers — models occasionally misattribute the cause when multiple similar code paths exist.
• Ask explicitly for “not just a try/catch” or you’ll sometimes get a fix that hides the symptom instead of solving the underlying issue.