Linux Documentation
 help / color / mirror / Atom feed
From: Krzysztof Kozlowski <krzk@kernel.org>
To: Sasha Levin <sashal@kernel.org>
Cc: Theodore Tso <tytso@mit.edu>,
	workflows@vger.kernel.org, ksummit@lists.linux.dev,
	linux-doc@vger.kernel.org, linux-kernel@vger.kernel.org,
	Jonathan Corbet <corbet@lwn.net>,
	Shuah Khan <skhan@linuxfoundation.org>,
	Randy Dunlap <rdunlap@infradead.org>,
	Mark Brown <broonie@kernel.org>,
	Laurent Pinchart <laurent.pinchart@ideasonboard.com>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Leon Romanovsky <leon@kernel.org>,
	Konstantin Ryabitsev <konstantin@linuxfoundation.org>,
	Thorsten Leemhuis <linux@leemhuis.info>,
	"Michael S. Tsirkin" <mst@redhat.com>,
	Kees Cook <kees@kernel.org>,
	Geert Uytterhoeven <geert@linux-m68k.org>,
	Dan Carpenter <error27@gmail.com>,
	Miguel Ojeda <ojeda@kernel.org>,
	Alexei Starovoitov <alexei.starovoitov@gmail.com>,
	Julia Lawall <julia.lawall@inria.fr>, SJ Park <sj@kernel.org>,
	Bradley Morgan <brads@mainlining.org>,
	stable@vger.kernel.org
Subject: Re: [PATCH v5 0/6] agents: add skills for Fixes tags, stable tags, similar bugs, and commit messages
Date: Thu, 12 Nov 2026 15:17:30 +0100	[thread overview]
Message-ID: <6af42f08-c052-48f4-b0e7-32cb340edbff@kernel.org> (raw)
In-Reply-To: <avXJSt_3DYx6aQsD@laps>

On 12/11/2026 15:12, Sasha Levin wrote:
> On Thu, Nov 12, 2026 at 02:49:09PM +0100, Krzysztof Kozlowski wrote:
>> On 12/11/2026 14:46, Theodore Tso wrote:
>>> On Thu, Nov 12, 2026 at 09:26:09AM -0500, Krzysztof Kozlowski wrote:
>>>>
>>>> All these are docs (even if you call them as a skill, it's the same),
>>>> therefore they should be placed in Documentation, not in top-level. We
>>>> should not create one documentation for humans and another for agents.
>>>> It's just one documentation.
> 
> Right. The idea here is that agent skills will mostly refer to our existing
> Documentation/ and will only act as a glue layer to make agents behave better.
> 
> Consider, for example, one of the most common feedback points we have to AI
> users: "make your commit message way less verbose". This is the type of things
> that would make sense as an agent-only instruction.

I am not objecting to have the instructions for agents. They should be
however placed in Documentation.
> 
>>> Skills aren't necessarily *just* documentation.  The skills
>>> specification[1] also defines a structured location in the skill's
>>> directory hierarchy for Python scripts or other executables.
>>
>> Specification is also documentation, just like for example specification
>> for our ABI. It's also located in Documentation directory.
> 
> One of the reasons to keep it out of Documentation/ was that it doesn't mesh
> with the infrastructure Documentation/ uses. For example, Documentation/ is all
> .rst, while the agents.md spec specifically calls out for markdown, and a very
> particular construction of that file.

Documentation already has YAML files for Devicetree and other ABI, thus
difference of format does not matter. It's just minor technical aspect -
tweak the Makefile.

> 
>> Agents supposedly "are just tools", thus just as sparse, smatch and
>> compilers such tools do not get another top-level directory.
> 
> I'm a bit confused by this... the original reasoning was the it doesn't really
> fit into either Documentation/ , scripts/, or tools/, so we just gave it it's

Oh, it can go to tools/, I don't mind.

> own home. Why does it matter if it's a top-level dir or not?

Because we do not need one more top-level directory for something which
is already categorized. We do not create another top-level for
"drivers-embedded" or "drivers-acpi". We do not have "tools" and
"tools-ai". Neither "Documentation" and
"Documentation-for-ai-but-we-will-call-it-agents".


> 
> It's something we can always move around later - agents don't care where it
> lives as long as it's refered to by the main AGENTS.md, so if we later come to
> the conclusion that it makes sense to move it, for example, under scripts/,
> then we can easily do that.

So place it deeper please.

Best regards,
Krzysztof

  reply	other threads:[~2026-11-12 14:17 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-11 22:59 [PATCH v5 0/6] agents: add skills for Fixes tags, stable tags, similar bugs, and commit messages Theodore Ts'o
2026-11-11 22:59 ` [PATCH v5 1/6] agents: add infrastructure for agent resources Theodore Ts'o
2026-11-11 22:59 ` [PATCH v5 2/6] agents: add a skill to determine Fixes tags Theodore Ts'o
2026-11-11 22:59 ` [PATCH v5 3/6] agents: add a skill to determine stable tags Theodore Ts'o
2026-11-12  1:10   ` Andrew Morton
2026-11-12  2:32     ` Sasha Levin
2026-11-12  2:50       ` Andrew Morton
2026-11-12  7:06   ` Miguel Ojeda
2026-11-11 22:59 ` [PATCH v5 4/6] agents: add a skill to find other instances of a fixed bug Theodore Ts'o
2026-11-11 22:59 ` [PATCH v5 5/6] agents: add a skill to write commit messages Theodore Ts'o
2026-11-11 22:59 ` [PATCH v5 6/6] agents: add a skill to write KUnit tests Theodore Ts'o
2026-11-12  0:57   ` Mauro Carvalho Chehab
2026-11-12  1:12   ` Andrew Morton
2026-11-12  8:26 ` [PATCH v5 0/6] agents: add skills for Fixes tags, stable tags, similar bugs, and commit messages Krzysztof Kozlowski
2026-11-12 13:46   ` Theodore Tso
2026-11-12 13:49     ` Krzysztof Kozlowski
2026-11-12 14:12       ` Sasha Levin
2026-11-12 14:17         ` Krzysztof Kozlowski [this message]
2026-11-12 14:42       ` Theodore Tso
2026-11-12 14:56         ` Krzysztof Kozlowski
2026-11-12 14:32 ` Sasha Levin

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=6af42f08-c052-48f4-b0e7-32cb340edbff@kernel.org \
    --to=krzk@kernel.org \
    --cc=alexei.starovoitov@gmail.com \
    --cc=brads@mainlining.org \
    --cc=broonie@kernel.org \
    --cc=corbet@lwn.net \
    --cc=error27@gmail.com \
    --cc=geert@linux-m68k.org \
    --cc=gregkh@linuxfoundation.org \
    --cc=julia.lawall@inria.fr \
    --cc=kees@kernel.org \
    --cc=konstantin@linuxfoundation.org \
    --cc=ksummit@lists.linux.dev \
    --cc=laurent.pinchart@ideasonboard.com \
    --cc=leon@kernel.org \
    --cc=linux-doc@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux@leemhuis.info \
    --cc=mst@redhat.com \
    --cc=ojeda@kernel.org \
    --cc=rdunlap@infradead.org \
    --cc=sashal@kernel.org \
    --cc=sj@kernel.org \
    --cc=skhan@linuxfoundation.org \
    --cc=stable@vger.kernel.org \
    --cc=tytso@mit.edu \
    --cc=workflows@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox