使い方
- インストールコマンドをコピーし、自分の環境へ導入する前にソースを確認します。
- 必要なツール、認証情報、対応環境を原文で確認します。
- 入力と成果物を指定し、出力を資料と照合します。
npx skills add https://github.com/olsonbd/superpowers-antigravity --skill implementing-schema-org開始前の確認
- 導入にはモデル契約、API枠、ファイルへのアクセス許可は含まれません。
- スター数や導入数は活動の指標であり、独立した品質・安全性試験ではありません。
見出し、例、コード、表、リンク、参照画像を含む原文を表示しています。
Implementing Schema.org Data Models
Overview
Schema.org is the internet's standard for structured data. Adopting it prevents "bikeshedding" over field names and ensures our data is ready for SEO, AI agents, and external integrations (like calendars) from day one.
Core Principle: If Schema.org has a standard for it, use it. Do not invent your own field names.
When to Use
- Creating a new Collection (e.g., Venues, Events, Users)
- Defining API responses
- Mapping data for export
The Rules
- Strict Naming: Use the exact property name from Schema.org (e.g.,
postalCodenotzip,givenNamenotfirstName). - Preserve Structure: If Schema.org defines a property as a nested Type (like
address->PostalAddress), you MUST nest it. Do not flatten it for convenience. - Use Standard Types: Map your entities to the most specific Schema.org Type available (e.g.,
Event->SocialEventif applicable).
Common Mappings
| Our Concept | Schema.org Type | Key Fields |
|---|---|---|
| Venue | Place | name, address (PostalAddress), geo (GeoCoordinates), publicAccess |
| Event | Event | name, startDate, endDate, location (Place), organizer (Person/Organization) |
| User | Person | givenName, familyName, email, telephone |
Implementation Pattern (Payload CMS)
❌ BAD: Flattened & Ad-Hoc
// Venues.ts
fields: [
{ name: "street", type: "text" }, // Violation: Flattened
{ name: "zip", type: "text" }, // Violation: Wrong name
{ name: "lat", type: "number" }, // Violation: Wrong name & flattened
];✅ GOOD: Nested & Standard
// Venues.ts
fields: [
{
name: "address", // Schema.org property
type: "group", // Preserves structure
fields: [
{ name: "streetAddress", type: "text" },
{ name: "addressLocality", type: "text" }, // City
{ name: "addressRegion", type: "text" }, // State
{ name: "postalCode", type: "text" }, // Zip
{ name: "addressCountry", type: "text" },
],
},
{
name: "geo",
type: "group",
fields: [
{ name: "latitude", type: "number" },
{ name: "longitude", type: "number" },
],
},
];JSON-LD Serialization
When exposing this data via API, simply wrap it to produce valid JSON-LD:
// Example response structure
{
"@context": "https://schema.org",
"@type": "Place",
"name": venue.name,
"address": {
"@type": "PostalAddress",
...venue.address // Spread the nested group directly!
},
"geo": {
"@type": "GeoCoordinates",
...venue.geo
}
}Benefit: Because we preserved the structure in the database, serialization is just a spread operator (...venue.address). No complex mapping code required.
Red Flags - STOP and Fix
- "I'll just call it
zipbecause it's shorter." -> STOP. UsepostalCode. - "Nested groups are annoying to query." -> STOP. Structure matters more than convenience.
- "I'll add a
customField." -> Check Schema.org first. Is there really no standard field?

