Re: AGENTS.md

On 2026-08-06 10:28, Cecilio Pardo wrote:
> On 8/6/2026 8:03 AM, Jean Louis wrote:
> 
>> I was all the time thinking that Emacs policy is not to accept LLM 
>> generated code.
>> 
>> Now you said it is the policy not to accept LLM-based agent generated 
>> code.
>> 
>> Please, clarify it, is it fine then to generate code by LLM if we do 
>> not use LLM-based agents?
> 
> You seem to know more than most about this topic, marking the 
> difference between "LLM generated" and "LLM agent generated" as a 
> significant one.
> 
> Will you please use that knowledge to submit a patch that does more and 
> the proposes AGENTS.md, and does it better, and does it without bowing 
> to "Big AI"?

I understand though OpenAI did much evil to the world, so as AGENTS.md 
is their "contribution" to AGENTIC AI FOUNDATION run by the spoiled 
Linux Foundation, I would contribute just nothing back to them.

AGENTS.md is and still remains a "deal" among proprietary companies. 
That other software packages adopted it means very little. Other 
software packages mostly use proprietary providers anyway.

It increases context length. That means money increase for each of those 
proprietary providers.

In Emacs it is easy to assign "memory" or additional context to any 
prompt or any file or any project.  Yet how is that helpful to them who 
run the SaaS LLMs? It is not.

They need AGENTS.md as that automatically increases context and their 
income flow.

Not yours.

The more detailed, the better. The more fat the AGENTS.md, so more money 
for them.

It may be ingested with every single prompt.

That is millions of dollars for them, proprietary companies dealing with 
proprietary software, controlling the world for their own selfish 
growth.

-- 
Jean Louis