Re: AGENTS.md

> Date: Wed, 05 Aug 2026 21:05:41 +0300
> From: Jean Louis <[email protected]>
> Cc: [email protected], [email protected], [email protected], [email protected]
> 
> On 2026-08-05 12:31, Sean Whitton wrote:
> > Richard Stallman [05/Aug  1:06am -04] wrote:
> >> 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?
> > 
> > I tested a commercial LLM at work and it obeyed the AGENTS.md and
> > refused to generate code.
> 
> And how could LLM look into the file?

Many do it automatically, which is why this file was added.

> So you did not test any LLM to obey the AGENTS.md but you tested some 
> software that uses LLMs.
> 
> LLM cannot see AGENTS.md unless you give it to it.

Many do it automatically.

> So you tested some commercial software that uses LLM inference to verify 
> if prompt extension is loaded... and? I am trying to understand the 
> point.

The point is stated above: "it obeyed the AGENTS.md and refused to
generate code".

> Did you test to ignore it?

In the case of LLMs that read such files automatically, ignoring it
would need an explicit action from the user instructing the agent to
ignore the file.  Such malevolent actions are out of scope for this
and other similar measures, which are intended to help users who act
in good faith.

This was explained to you many times, and you still keep reiterating
the now-irrelevant questions and information.

> Did you test all the millions of LLMs on it?

Is this a serious question?

> Or we are just creating AGENTS.md for commercial software using LLM?!

Again, we created it to help users making innocent mistakes, using
whatever LLM-based agents, both proprietary and free.  The Emacs
policy is not to accept code generated by LLM-based agents regardless
of whether they are free or not.  As was repeatedly explained here.