forked from catchorg/Catch2
Add LLM policy to contributing.md
This commit is contained in:
@@ -7,6 +7,8 @@
|
||||
[Writing documentation](#writing-documentation)<br>
|
||||
[Writing code](#writing-code)<br>
|
||||
[CoC](#coc)<br>
|
||||
[Using LLMs when contributing](#using-llms-when-contributing)<br>
|
||||
|
||||
|
||||
So you want to contribute something to Catch2? That's great! Whether it's
|
||||
a bug fix, a new feature, support for additional compilers - or just
|
||||
@@ -332,6 +334,19 @@ When adding new `CATCH_CONFIG` option, there are multiple places to edit:
|
||||
This project has a [CoC](../CODE_OF_CONDUCT.md). Please adhere to it
|
||||
while contributing to Catch2.
|
||||
|
||||
|
||||
## Using LLMs when contributing
|
||||
|
||||
I do not care whether you used LLM for your contribution. What I care
|
||||
about is the quality of the contribution, not whether you made it through
|
||||
prompting LLM, letting GAs run wild, dictated it into the computer, wrote
|
||||
it using your nose or used butterflies to flip the correct bits.
|
||||
|
||||
The flipside of this is that I am also not going to iterate your LLM for
|
||||
you. If a PR looks LLM generated and does not pass the muster to be merged
|
||||
as-is, I am going to close it.
|
||||
|
||||
|
||||
-----------
|
||||
|
||||
_This documentation will always be in-progress as new information comes
|
||||
|
||||
Reference in New Issue
Block a user