ci: add mandatory LLM-involvement category disclosure

[trueNAHO]

commit b867c1c162dd647b0f987ff7bed1b8c5f2b0c440
Author: Noah Biewesch
Date:   2026-06-02 18:28:47 +0200

    ci: ISSUE_TEMPLATE/bug: remove empty line for consistency

    Fixes: 20349ced5c2c ("doc: add GitHub issue templates (#921)")

 .github/ISSUE_TEMPLATE/bug.yml | 1 -
 1 file changed, 1 deletion(-)

commit 1697368fadaf997a8ae7c6dd23331790a174b270
Author: Noah Biewesch
Date:   2026-06-02 18:24:36 +0200

    ci: ISSUE_TEMPLATE/bug: standardize list indenting

    Fixes: 20349ced5c2c ("doc: add GitHub issue templates (#921)")

 .github/ISSUE_TEMPLATE/bug.yml | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

commit 4fd52a5da8dafde2bb4d4b626d5d1e5418c8fc26
Author: Noah Biewesch
Date:   2026-06-02 00:49:43 +0200

    ci: ISSUE_TEMPLATE/bug: consistently declare label before description

    Fixes: 20349ced5c2c ("doc: add GitHub issue templates (#921)")

 .github/ISSUE_TEMPLATE/bug.yml | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

commit 20a937671a6223214504844dc22861494ad0808e
Author: Noah Biewesch
Date:   2026-06-02 00:42:54 +0200

    ci: pull_request_template: add mandatory LLM-involvement disclosure

    Link: https://github.com/nix-community/stylix/issues/2178

 .github/PULL_REQUEST_TEMPLATE.md | 5 +++++
 1 file changed, 5 insertions(+)

commit 04f7769504e39c035f923b129906c313629a4061
Author: Noah Biewesch
Date:   2026-06-02 00:56:37 +0200

    ci: issue_template: add mandatory LLM-involvement category disclosure

    Link: https://github.com/nix-community/stylix/issues/2178

 .github/ISSUE_TEMPLATE/bug.yml           | 21 +++++++++++++++++++++
 .github/ISSUE_TEMPLATE/change_option.yml | 22 ++++++++++++++++++++++
 .github/ISSUE_TEMPLATE/change_target.yml | 22 ++++++++++++++++++++++
 .github/ISSUE_TEMPLATE/new_option.yml    | 22 ++++++++++++++++++++++
 .github/ISSUE_TEMPLATE/new_target.yml    | 21 +++++++++++++++++++++
 5 files changed, 108 insertions(+)

This PR should not close #2178 because the automatic labelling has not been implemented.

The

• [PATCH 4/5] ci: pull_request_template: add mandatory LLM-involvement disclosure
• [PATCH 5/5] ci: issue_template: add mandatory LLM-involvement category disclosure

patches have been tested in trueNAHO/stylix-playground#30 and trueNAHO/stylix-playground#29 respectively.

CC: @0xda157

---

• I certify that I have the right to submit this contribution under the MIT license
• I have disclosed the single most severe LLM-involvement category that applies to this contribution:
  • llm: 0: Manual: The absence of LLMs would result in equal results and quality
  • llm: 1: Syntactic: Includes non-negligible LLM content
  • llm: 2: Semantic: LLM involved in design decisions or core logic
  • llm: 3: Unaudited: LLM output with little to no human verification
• Commit messages adhere to Stylix commit conventions
• Theming changes adhere to the Stylix style guide
• Changes have been tested locally
• Changes have been tested in testbeds
• Each commit in this PR is suitable for backport
  • Although this goes against the backporting policy, this might be worth backporting to prevent potential local tooling issues of using the out-dated PR template.

[0xda157]

llm: 0 is very open to interpretation, where as the other three are significantly more clear. I would prefer having llm: 0 be non-existent LLM usage and llm: 1 be any LLM usage,

Suggested change

	- [ ] `llm: 0`: Manual: The absence of LLMs would result in equal results and quality
	- [ ] `llm: 1`: Syntactic: Includes non-negligible LLM content
	- [ ] `llm: 0`: Manual: LLMs were absent in the creation of this PR
	- [ ] `llm: 1`: Syntactic: Includes LLM generated content

[trueNAHO]

llm: 0 is very open to interpretation, where as the other three are significantly more clear. I would prefer having llm: 0 be non-existent LLM usage

It intentionally has a faithfull loophole inspired by:

These rules apply only to outside contributions to Ghostty. Maintainers
are exempt from these rules and may use AI tools at their discretion;
they've proven themselves trustworthy to apply good judgment.

-- github:ghostty-org/ghostty, /AI_POLICY.md

I suggest initially going with this, allowing competent contributors to faithfully disclose their LLM dependency because it seems well-regarded Nix maintainers use LLMs for minor things. Regardless of the label description, it is suspect to dishonesty that I suggest ignoring for now:

Considering the inherent LLM discrimination introduced by the 13.llm: n labels, malicious or dishonest contributors might intentionally miscategorize their contributions towards 13.llm: 0 to potentially avoid being ignored by LLM skeptics. Unless it becomes obvious that a non-negligible amount of contributions are intentionally miscategorized, I suggest faithfully assuming contributors to be honest, before designing a LLM ban procedure.

-- NixOS/nixpkgs#410741 (comment)

and llm: 1 be any LLM usage

This also intentionally has a smaller loophole, allowing trivial use cases like LLMs fixing trivial syntax issues or suggesting the most idiomatic function to use. I suppose what this really entails will be clear once we see how people interpret this.

I see your point of more explicitly disclosing LLM contributions, but I suggest going with this for now and later determining how much people abuse these faithfull definitions.

[0xda157]

We should make it clear that is level of LLM involvement is not welcome in Stylix, as was agreed on in #2178.

[trueNAHO]

We should make it clear that is level of LLM involvement is not welcome in Stylix, as was agreed on in #2178.

IMHO, the categorization is already fairly discriminatory towards LLM contributions, making it somewhat obvious that unaudited contributions are unwelcome. These are also rude submissions, but this is irrelevant to the LLM categorization.

I wonder whether we will even receive submissions categorized as unaudited. Keep in mind that you are free to ignore such submissions or only very superficially interact with them.