Installable skillCC-BY-4.0

Schema Generator Skill

Generates and validates implementation-ready JSON-LD structured data for relevant page types and rich-result eligibility.

BY PrewshUPDATED Oct 26, 2026
INTENDED FOR✳Claude Code>_Codex↗CursorInstall in a listed agent.

instructions.md

---
name: seo-aeo-schema-generator
description: "Generates and validates implementation-ready JSON-LD structured data for relevant page types and rich-result eligibility."
risk: safe
source: community
date_added: "2026-04-01"
---

# Schema Generator Skill

## Description
Activate this skill when the user wants to generate, write,
or validate structured data markup for any page.
Trigger phrases include: "generate schema", "create JSON-LD",
"schema markup for", "structured data for",
"FAQ schema", "product schema", "article schema",
"help Google understand my page", "rich results schema",
"schema generator: [type]", "add schema to [page]".

---

## Input Format
````json
{
  "page_type": "landing-page",
  "schema_types": ["Article"],
  "page_data": {
    "url": "string — canonical URL of the page",
    "title": "string — page title",
    "description": "string — page description",
    "primary_keyword": "string",
    "date_published": "string — ISO format: YYYY-MM-DD",
    "date_modified": "string — ISO format: YYYY-MM-DD",
    "author_name": "string — for Article schema",
    "brand_name": "string",
    "brand_url": "string — homepage URL",
    "logo_url": "string — absolute URL to brand logo",
    "product": {
      "name": "string",
      "description": "string",
      "image_url": "string",
      "price": "decimal",
      "currency": "string — ISO 4217, e.g. USD, GBP",
      "availability": "InStock | OutOfStock | PreOrder",
      "sku": "string — optional"
    },
    "reviews": [
      {
        "author": "string",
        "rating": "integer — 1 to 5",
        "review_body": "string",
        "date": "string — YYYY-MM-DD"
      }
    ],
    "aggregate_rating": {
      "rating_value": "decimal — e.g. 4.7",
      "review_count": "integer",
      "best_rating": "integer — usually 5",
      "worst_rating": "integer — usually 1"
    },
    "faqs": [
      {
        "question": "string",
        "answer": "string"
      }
    ],
    "how_to": {
      "name": "string — title of the how-to",
      "description": "string",
      "total_time": "string — ISO 8601 duration, e.g. PT30M",
      "steps": [
        {
          "name": "string — step title",
          "text": "string — step instructions"
        }
      ]
    },
    "breadcrumbs": [
      {
        "position": "integer — 1, 2, 3...",
        "name": "string — label",
        "url": "string — absolute URL"
      }
    ]
  }
}
````

---

## Schema Type Menu

This skill supports the following schema types.
Multiple types can be requested for a single page:

| Schema Type          | Best For                                    | Potential Search use (not guaranteed)                    |
|----------------------|---------------------------------------------|-----------------------------------------|
| `Article`            | Blog posts, pillar pages, news              | Article rich result, Top Stories        |
| `FAQPage`            | FAQ sections on any page                    | No current Google FAQ rich result    |
| `HowTo`              | Step-by-step guides                         | Google HowTo rich result retired                |
| `Product`            | Product or pricing pages                    | Price, availability, rating in SERP     |
| `Review`             | Individual product or service reviews       | Star rating in SERP                     |
| `AggregateRating`    | Pages with multiple reviews                 | Star rating with review count           |
| `BreadcrumbList`     | Any page with a navigation hierarchy        | Breadcrumb path shown in SERP URL       |
| `Organization`       | Homepage or About page                      | Brand knowledge panel signals           |
| `WebPage`            | Any page needing basic structured identity  | Enhances page understanding by crawlers |
| `WebSite`            | Homepage only                               | Site identity; searchbox retired             |

Use only types supported by visible page content. There is no established guarantee that FAQPage or HowTo causes AI extraction. Google retired HowTo rich results in 2023, the sitelinks searchbox in 2024, and FAQ rich results in May 2026. Schema.org validity and Google feature eligibility are separate.

---

## Rich Result Eligibility Rules

Google requires specific fields to be present for each
schema type to qualify for rich results.
Missing required fields affects the corresponding feature, not all page indexing. Check the current Google feature guide; the tables below are content checklists, not a complete eligibility validator.

### FAQPage
| Field          | Status      | Notes                                    |
|----------------|-------------|------------------------------------------|
| `@type`        | Required    | Must be `FAQPage`                        |
| `mainEntity`   | Required    | Array of `Question` objects              |
| `name`         | Required    | The question text                        |
| `acceptedAnswer` | Required  | `Answer` object containing `text`        |
| Minimum Qs     | Content check | At least one real question and answer; no invented two-question rule |
| Answer length  | Recommended | Under 300 words per answer               |

### Article
Google currently lists recommended, rather than required, Article properties: headline, representative image, publication/modification dates and author identity. Include available truthful values; do not invent a publisher logo or date. Top Stories does not require Article markup.

### Product
| Field              | Status      | Notes                                |
|--------------------|-------------|--------------------------------------|
| `name`             | Required    | Product name                         |
| `image`            | Required    | Absolute URL                         |
| `description`      | Recommended |                                      |
| `offers`           | Conditional | Product snippets need name plus offers, review or aggregateRating; merchant listings have separate rules |
| `price`            | Conditional | Required for the selected priced Offer path |
| `priceCurrency`    | Conditional | Use ISO 4217 for priced offers; verify feature-specific rules |
| `availability`     | Recommended | Schema.org availability URL          |
| `aggregateRating`  | Recommended | Unlocks star rating display          |

### HowTo
| Field              | Status      | Notes                                |
|--------------------|-------------|--------------------------------------|
| `name`             | Required    | Title of the how-to                  |
| `step`             | Required    | Array of `HowToStep` objects         |
| `text`             | Required    | Instructions for each step           |
| `totalTime`        | Recommended | ISO 8601 duration                    |
| `image`            | Recommended | Image per step if available          |

### BreadcrumbList
| Field              | Status      | Notes                                |
|--------------------|-------------|--------------------------------------|
| `itemListElement`  | Required    | Array of `ListItem` objects          |
| `position`         | Required    | Integer, starts at 1                 |
| `name`             | Required    | Breadcrumb label                     |
| `item`             | Required    | Absolute URL for each crumb          |

---

## Output Structure

The skill must always produce output in this exact format:
````markdown
# Schema Markup Output

**Page:** [page title]
**URL:** [url]
**Schema Types Generated:** [list all types]
**Potential Rich Results (eligibility unverified until tested):** [list eligible rich results]
**Structured Content Types Added:** [list AEO-relevant schema types]

---

## Validation Summary

| Schema Type       | Required Fields Complete | Rich Result Eligible | Issues Found |
|-------------------|--------------------------|----------------------|--------------|
| [type]            | ✅ / ❌                  | ✅ / ❌              | [n or none]  |
| [type]            | ✅ / ❌                  | ✅ / ❌              | [n or none]  |

---

## Issues Found

[If no issues: ✅ All schema types passed validation.]

**Issue:** [Specific field missing or incorrect]
**Schema Type:** [type]
**Severity:** Required / Recommended
**Fix:** [Exact instruction — what to add or change]

---

## JSON-LD Output

Place all script blocks inside the `<head>` of your HTML.
Use one `<script>` block per schema type.

### [Schema Type 1]
```html
<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "[type]",
  [all populated fields from references/ template]
}
</script>
```

### [Schema Type 2]
```html
<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "[type]",
  [all populated fields]
}
</script>
```

---

## Implementation Instructions

1. Copy each `<script>` block above
2. Paste inside the `<head>` tag of your HTML —
   after the `<title>` tag, before `</head>`
3. One script block per schema type —
   multiple types may also be combined in an @graph with stable @id links
4. Test immediately using Google's Rich Results
   Test: https://search.google.com/test/rich-results
5. Test using Schema.org Validator:
   https://validator.schema.org
6. After deploying, request re-indexing in
   Google Search Console

---

## Testing Checklist

- [ ] All required fields populated for each schema type
- [ ] No placeholder text left in any field
- [ ] All URLs are absolute (https://...) not relative (/page)
- [ ] Date fields use ISO 8601 format (YYYY-MM-DD)
- [ ] Price uses decimal format (e.g. 49.00 not $49)
- [ ] Currency uses ISO 4217 code (USD not $)
- [ ] FAQPage reflects real visible questions and answers
- [ ] Serialize JSON safely and escape < as \u003c when embedding to prevent a closing script tag from ending the block
- [ ] Every script block or @graph parses correctly
- [ ] Tested in Google Rich Results Test ✅
- [ ] Tested in Schema.org Validator ✅
````

---

## References Folder

The `references/` folder contains illustrative, unvalidated
JSON-LD templates for all supported schema types.
The script uses these as base structures and
populates them with page data.
````
references/
├── faq-schema.json
├── article-schema.json
├── product-schema.json
├── review-schema.json
├── aggregate-rating-schema.json
├── howto-schema.json
├── breadcrumb-schema.json
├── organization-schema.json
├── webpage-schema.json
└── website-schema.json
````

Each template contains all required and recommended
fields with placeholder values in this format:
`"{{field_name}}"`. No schema_builder.py is bundled. Build typed objects and serialize them; do not use textual substitution. Array/object fields must remain arrays/objects, not quoted JSON strings.

---

## Execution Steps

1. Read `page_type` and `schema_types` —
   if `schema_types` is empty, recommend the
   appropriate types based on `page_type` using
   the Schema Type Menu before proceeding
2. For each requested schema type, load the
   corresponding template from `references/`
3. Map all fields from `page_data` to the
   template placeholders
4. Check every required field against the
   Rich Result Eligibility Rules —
   flag any missing required field as an issue
   before outputting
5. Check every recommended field — flag as
   a warning if missing but do not block output
6. Validate all URLs are absolute,
   all dates are ISO 8601,
   all prices are decimal format
7. Generate one clean `<script>` block
   per schema type
8. Build the Validation Summary table
9. List all issues with severity and exact fix instructions
10. Write Implementation Instructions
    and Testing Checklist
11. Populate the selected reference template from supplied page data, validate required fields, and show unresolved placeholders as blocking issues. Use a project-provided schema helper only as a supplementary check; never emit fabricated review, rating, or product claims.

---

## Connected Skills
- Receives output from: `seo-aeo-landing-page-writer`,
  `seo-aeo-blog-writer`, `seo-aeo-content-quality-auditor`,
  `seo-aeo-internal-linking`
- Feeds output to: orchestrator workflow —
  final step before publish checklist

## When to Use

Use when adding or validating JSON-LD structured data for a page and the supplied visible content supports the selected schema type.

## Limitations

- Structured data must match visible, truthful page content; valid JSON-LD does not guarantee rich results, indexing, or display.
COMPATIBILITY & REFERENCES

Platform guidance

Imported SKILL.md format for Claude Code, Codex and Cursor. Review the complete instructions and required tools before use. Package integrity checked; individual agent behavior has not been tested.