Re: AGENTS.md

> Date: Wed, 05 Aug 2026 12:48:22 +0300
> From: Jean Louis <[email protected]>
> CC: [email protected], [email protected], [email protected]
> 
> On August 5, 2026 8:06:42 AM GMT+03:00, Richard Stallman <[email protected]> wrote:
> 
> >Eli wrote:
> >
> >    Do we have any idea how effective is this in practice?  Does this
> >    prevent what we want to avoid with likelihood that is close enough to
> >    100% for us to consider it the "solution"?  Because if not, we'd need
> >    this _and_ the hooks as Plan B.
> 
> It's notion (agents.md)to have one file that modifies LLM prompt so that proprietary companies don't need to invent many different files so that their software can run easy and that they minimize their development cost.
> 
> It's not effective in practice.

This is false, as you have been pointed out already (and said then you
didn't know it).  Please be fair and mention this when you answer such
questions, instead of misleading people.  Here's that URL again:

  https://agents.md/

> User can simply say to software to ignore the agents.md file.
> 
> User can delete the file.

User can do anything, including copy someone else's code into his
contribution.  As was already explained to you, this file is for
helping people avoid inadvertent mistakes of generating code by
LLM-based tools.

> Too many software using LLM including Emacs packages don't observe agents.md

Many do, so this file is an improvement.

> >That is a good question.  Do we actually know how much, or in what way, LLMs'
> >behavior will be affected by an AGENTS.md file?
> 
> No, we do not know how LLMs will behave when they encounter AGENTS.md. We cannot predict it because:
> 
> · LLMs are pattern matchers, not intelligences
> · They don't "understand" meaning or intent
> · Their behavior depends on training data, model version, and user prompting
> · Different models respond differently to the same prompt
> · The same model responds differently at different times

Again, this is misleading, because evidence to the contrary was
already posted here.  Please don't spread misinformation.

> We are guessing that AGENTS.md will influence LLMs.

No, we are not guessing.  The effects of this file were tested.  Once
again, this was reported here, so you again spread misinformation.

> Users who wish to generate LLM code cannot be prevented by it.

Indeed, which is why this file is not for those cases.  We plan to
deploy Git hooks for that.