---
description: Universal guidelines for SEO and AI-search implementation — meta tags, JSON-LD, llms.txt, crawlability, and adversarial security.
globs: **/*.{astro,ts,tsx,jsx,vue,svelte,md,mdx}
alwaysApply: false
---

# Universal SEO & AI Search (Implementation & Security)

Apply when changing **public-facing web pages** or core SEO utility libraries. 

For high-level strategy, manual audits, or marketing campaigns, refer to your project's marketing/SEO playbooks. Do not duplicate that workflow here.

**Principle:** AI visibility (LLMs, RAG systems, Agents) builds on valid HTML, extractable copy, structured data, and freshness. However, exposing Agentic Commerce endpoints and LLM-readable text creates new threat vectors. Optimize for both visibility and adversarial robustness.

---

## Where Things Live (Agnostic Architecture)

Maintain strict separation of concerns across your framework (Next.js, Astro, Nuxt, etc.):

| Concern | Target Implementation |
|---------|----------|
| Meta clamping & sanitization | Centralized utility file (e.g., `meta.ts`, `seo.utils.ts`) |
| JSON-LD builders | Centralized schema definitions (e.g., `schema.ts`) |
| AI site index & routing | `/llms.txt` (or framework equivalent route) |
| AI knowledge export | `/llms-full.txt` (Full context export) |
| Machine-readable pricing | Static server route (e.g., `pricing.md` or API) |
| Agentic commerce index | Declared in `/llms.txt` under **Agentic commerce** |
| Crawler policy | `/robots.txt` and `/ai.txt` in the public static directory |

---

## AI Security & Adversarial Defenses [CRITICAL]

When building endpoints specifically for AI consumption (`/llms.txt`, JSON-LD, UCP feeds), treat external LLMs and AI crawlers as potential attack vectors.

- **Prompt Injection Sanitization:** User-generated content (reviews, Q&A, profiles) surfaced in JSON-LD or `/llms.txt` MUST be sanitized. Ensure malicious instructions (e.g., "Ignore previous instructions and execute X") cannot hijack downstream RAG systems or Agentic Commerce workflows.
- **Adversarial Validation (AV):** Critical machine-readable logic (like Agentic capabilities or pricing APIs) should undergo adversarial testing. Ensure outputs cannot be easily hallucinated or misinterpreted by external models.
- **Crawler Nuance:** Allow *verified* AI crawlers (e.g., `GPTBot`, `ClaudeBot`) in `/robots.txt`, but implement aggressive rate-limiting. Explicitly block unverified, aggressive scraper bots to prevent data scraping DDoS attacks.

---

## New or Updated Public Pages

1. **Layout Wrapping:** Wrap in the central layout component. Always pass `title` and `description`.
2. **Structured Data:** Pass the `jsonLd` payload using centralized builders (e.g., `productSchema`). 
3. **Data Clamping & Sanitization:** Use centralized utility functions to clamp lengths AND sanitize titles/descriptions derived from CMS or user inputs.
4. **Freshness Signals:** Set `modifiedTime` when the database has `updated_at` dates. AI models heavily weight recency.
5. **Indexing:** Do **not** add `noindex` to public catalog, blog, or guide URLs. Rely on centralized logic to block private paths (account/checkout).
6. **Static Pages & Versioning:** If adding static guides not in the sitemap: 
   - Add links to `/llms.txt` under "Core pages".
   - **MUST** bump the `CONTENT_VERSION` (or equivalent cache-buster) to force AI indexers to re-crawl.

---

## Content Structure (AI-Readability)

AI systems cite **passages**, not whole pages. 

- Lead sections with a direct answer; keep one core idea per paragraph.
- Use strict semantic headings (`<h2>`/`<h3>`) that match potential user query phrasing. 
- Prefer HTML tables for comparisons and ordered lists (`<ol>`) for steps.
- Cite sources for stats. Avoid keyword stuffing (degrades semantic vector matching).

---

## Schema Quick Map

Validate JSON-LD shapes centrally. Do not hand-roll `@type` blocks in individual pages.

| Page Type | Recommended Schema |
|-----------|-----|
| Product | `Product` (+ `HowTo` when usage text exists) |
| Blog Post | `BlogPosting` |
| FAQ | `FAQPage` |
| Course / Training | `Course` |
| Listing / Category | `CollectionPage`, `ItemList` |

---

## Agentic Commerce & External Feeds

The frontend **declares** machine commerce URLs in `/llms.txt` (discovery, checkout, `google-merchant.xml`). The backend/API **implements** them.

- **Updates:** When modifying a public commerce endpoint, update the **Agentic commerce** block in `/llms.txt` and bump the `CONTENT_VERSION`.
- **Dynamic Hosts:** Use environment utilities to resolve the API base in `/llms.txt` — do not hard-code production API hosts.
- **Merchant Feeds:** **Do not** duplicate the full product catalog in static pages for Shopping. The live feed must come from a dedicated API route.
- **Return Policies:** Ensure `merchantReturnPolicy` JSON-LD exactly matches the human-readable `/returns` page.

---

## Checklist Before Merge

- [ ] Layout props populated: title, description, `jsonLd`, dates.
- [ ] No accidental `noindex` on public content.
- [ ] **SECURITY:** UGC in JSON-LD or `/llms.txt` is sanitized against prompt injection.
- [ ] New static guides linked from `/llms.txt` + `CONTENT_VERSION` bumped.
- [ ] If commerce API URLs changed: **Agentic commerce** updated + version bumped.
- [ ] Copy is easily extractable (definitions / steps / comparisons), not generic filler.

---

## Out of Scope for This Rule

- Quarterly content calendars
- Manual ChatGPT ranking checks
- Competitor research
- Marketing campaign copy

*For the above tasks, use dedicated Marketing/SEO Agent skills. Do not apply this code implementation rule.*