FR: Support for `llms.txt`

[Golembivskyi]

Hello Ether Creative team!

I'm a big fan of your SEO plugin — it's been a game-changer for managing meta tags, sitemaps, and redirects on my Craft CMS sites. The dynamic robots.txt rendering from a Twig template is especially handy, and I've been using it to keep things flexible without hardcoding paths.

Feature Request: Support for llms.txt

With the rise of AI crawlers (like GPTBot, Claude, Gemini, and PerplexityBot) in 2025, there's a new standard emerging: llms.txt (see llms.txt.org). It's essentially a robots.txt-like file specifically for AI/LLM bots, allowing site owners to control what gets indexed for training data or responses. It uses the same syntax (User-agent, Allow, Disallow, etc.) but focuses on AI-specific policies, like attribution requirements or quoting limits.

Right now, users like me have to manually create and maintain a static llms.txt file in the root (or /.well-known/llms.txt), which feels outdated in a CMS like Craft. Since your plugin already handles dynamic robots.txt via Twig templates, it would be amazing to extend this to llms.txt!

Proposed Implementation

• Dynamic Rendering: Add a new setting in SEO → Settings (e.g., "LLMs" section) where users can configure a Twig template path for llms.txt, similar to the existing Robots setting.
  • Default template: Something like seo/_seo/llms (with access to seo object for variables like site URLs, disallowed paths, etc.).
  • Serve it at /llms.txt (and optionally /.well-known/llms.txt for better compatibility).
• Built-in Defaults: Include a sample template with common directives, e.g.: User-agent: * Allow: / Disallow: /admin/

This could pull from Craft globals or SEO settings (e.g., exclude paths from sitemap exclusions).

• Integration Ideas:
• Link it to existing features: Auto-disallow paths from Redirects or Sitemap exclusions.
• Add a simple policy editor in the field type for quick attribution rules (e.g., "Require link to golem.agency").
• Support for multi-environment: Like robots.txt, render differently for dev/staging (e.g., full Disallow in non-production).

Why This Matters

• SEO for AI Era: More traffic comes from AI responses (ChatGPT, Gemini searches). Controlling llms.txt ensures content is indexed ethically — e.g., forcing attribution back to our sites, which boosts brand mentions.
• Craft Community Fit: Plugins like SEOmatic already touch on robots.txt; this would position Ether SEO as forward-thinking for 2025+.
• Low Effort, High Value: It's a small extension of your existing robots.txt logic — mostly routing and templating.

I'd be happy to test a beta or contribute a PR if you point me in the right direction! What do you think?

Thanks for the great work — keep crafting! 🚀

[alexjcollins]

Hello @Golembivskyi, Thanks very much for the thoughtful and detailed suggestion. We're also increasingly seeing the need for this feature in our agency work, and as such, we had already included it on our roadmap. When we will get to it, I am not sure. There is a possibility we might be able to implement it before the end of the year, but if not, it will be in Q1 26. Thanks.

[glenwith1n]

Would love this option within the SEO plugin, another tab for LLMS as we are currently doubling up on fields for how we are doing it.