api-resource Skill
Creates or modifies API Platform resources using DTOs, Object Mappers, and state providers or processors.
- GitHub Stars
- 29
- Dateien
- 1
- Letzter Commit
- 22. Okt. 2026
- Lizenz
- —
- Erlaubte Tools
- —
Was macht der api-resource Skill?
api-resource ist ein Agent Skill. Creates or modifies API Platform resources using DTOs, Object Mappers, and state providers or processors.
Zusammenfassung erzeugt aus der SKILL.md des Skills.
Über den api-resource skill
Developers building web APIs can use this skill to design, create, or modify API Platform resources. It assists in defining the public shape of data through resource declarations, ensuring that Hydra, OpenAPI, and GraphQL documentation remains accurate. The skill supports various architectural patterns, ranging from simple entity-as-resource setups for rapid prototyping to decoupled DTO strategies for complex systems.
When an agent needs to expose an entity over HTTP or reshape an existing data model, it can implement DTOs with the Object Mapper. This process uses stateOptions and the Map attribute to connect public representations to underlying Doctrine entities or Laravel Eloquent models. The skill also handles advanced requirements like configuring nested sub-resources with uriVariables, implementing custom state providers and processors for non-CRUD logic, and managing high-precision numbers using BcMath. It provides the necessary structure to handle input DTOs per operation, allowing different payloads for reading and writing data.
Was ein agent damit machen kann
- 1Expose an entity as an API resourceAn agent can quickly turn a database entity into a functional API endpoint for basic CRUD operations.
- 2Decouple API shape from database schemaDevelopers can use DTOs and an Object Mapper to ensure the public API contract does not change when the underlying database structure changes.
- 3Implement complex data transformationsYou can create custom transformers to handle intricate mapping logic between your data models and your API resources.
- 4Configure nested resource hierarchiesAn agent can set up complex URL structures for resources that belong to a parent entity, such as messages within a mailbox.
Geschrieben basierend auf dem SKILL.md des skills. Jeder use case wurde gegen ein zitat daraus geprüft.
So installierst du den api-resource Skill
Jeder Befehl kopiert den Skill-Ordner in das Skills-Verzeichnis des Clients. Starte den Client neu oder beginne eine neue Session, um ihn zu laden.Den api-resource Skill in Plugin installieren
/plugin marketplace add api-platform/skillset /plugin install api-platform@api-platform-skillset
Den api-resource Skill in Claude Code installieren
~/.claude/skills/api-resource/SKILL.md
$ npx degit api-platform/skillset/skills/api-resource ~/.claude/skills/api-resourceDen api-resource Skill in Claude Code (project) installieren
.claude/skills/api-resource/SKILL.md
$ npx degit api-platform/skillset/skills/api-resource .claude/skills/api-resourceDen api-resource Skill in Codex installieren
~/.agents/skills/api-resource/SKILL.md
$ npx degit api-platform/skillset/skills/api-resource ~/.agents/skills/api-resourceDen api-resource Skill in Gemini CLI installieren
~/.gemini/skills/api-resource/SKILL.md
$ npx degit api-platform/skillset/skills/api-resource ~/.gemini/skills/api-resourceDen api-resource Skill in Claude.ai installieren
npx degit api-platform/skillset/skills/api-resource api-resource && zip -r api-resource.zip api-resource # claude.ai → Customize → Skills → + → Create skill → Upload a skill → api-resource.zip
SKILL.md
Auf GitHub ansehenCreating API Platform Resources
Design-First Principle
An API Platform resource is built from three concerns (see https://api-platform.com/docs/core/design/):
- Resource declaration — a plain PHP object marked
#[ApiResource]describing the public shape. Single source of truth for Hydra, OpenAPI and GraphQL. - Data retrieval — a state provider hydrates that object (Get, GetCollection).
- Data persistence — a state processor writes it (Post, Put, Patch, Delete).
Design the public shape first; the resource class doesn't have to be a Doctrine entity. How you wire it to persistence is a separate decision:
| Approach | When | Hookup |
|---|---|---|
| Entity as Resource | Prototyping, plain CRUD, internal model == public shape | #[ApiResource] on the entity; built-in Doctrine provider/processor (zero wiring) |
| DTO with Object Mapper | Decoupled public shape over a Doctrine entity/document — whether fields match 1:1 or need renames/transforms | stateOptions: new Options(entityClass:/documentClass:) and #[Map] on the DTO |
| Input DTOs per Operation | Write model differs from read model (stricter create/update payloads) | per-operation input: + processor |
| Custom Provider/Processor | Non-CRUD, external data, complex domain logic, CQRS | hand-written ProviderInterface/ProcessorInterface (see state-provider/state-processor) |
Rule of thumb: entity-as-resource is convenient but couples your public contract to your schema — fine for prototypes, probably not for large or non-CRUD systems. Decouple with a DTO as soon as the two shapes diverge.
Entity as Resource (Simplest)
Mark the entity itself — built-in Doctrine providers/processors handle everything. No provider, processor, or mapping to write:
use ApiPlatform\Metadata\ApiResource;
use Doctrine\ORM\Mapping as ORM;
#[ORM\Entity]
#[ApiResource]
class Book
{
#[ORM\Id, ORM\GeneratedValue, ORM\Column]
public ?int $id = null;
#[ORM\Column]
public string $title = '';
}
Use this for prototypes and straight CRUD. Migrate to a DTO (strategies below) once the API shape must differ from the schema.
DTO with Object Mapper (Recommended)
One mechanism, not two: API Platform's ObjectMapperProvider /
ObjectMapperProcessor activate only when both are present — stateOptions
naming the entityClass: (or documentClass:) and a #[Map] attribute on the
DTO (and on the input entity for writes). On read it runs
map($entity, $resourceClass); on write map($inputDto, $entityClass) then persists
via the Doctrine processor. There is no Object-Mapper mode without stateOptions.
Fields that line up by name map automatically; add #[Map(source: …)] to rename or
transform: to convert. The same config covers both the trivial 1:1 case and
arbitrary renames/transforms:
<?php
namespace App\ApiResource;
use ApiPlatform\Doctrine\Odm\State\Options;
use ApiPlatform\Metadata\ApiResource;
use ApiPlatform\Metadata\Get;
use ApiPlatform\Metadata\GetCollection;
use App\Document\Order as DocumentOrder;
use App\Transformer\CustomerTransformer;
use Symfony\Component\ObjectMapper\Attribute\Map;
#[ApiResource(
operations: [
new Get(stateOptions: new Options(documentClass: DocumentOrder::class)),
new GetCollection(stateOptions: new Options(documentClass: DocumentOrder::class)),
]
)]
#[Map(target: DocumentOrder::class)]
final class Order
{
public string $id;
// Simple 1:1 field mapping (automatic when names match)
public string $status;
// Map from a different source field
#[Map(source: 'customerName')]
public string $buyer;
// Transform with a custom callable
#[Map(source: 'rawData', transform: new CustomerTransformer())]
public CustomerDto $customer;
}
Custom Transformers
Implement TransformCallableInterface for complex mappings:
<?php
namespace App\Transformer;
use Symfony\Component\DependencyInjection\Attribute\Exclude;
use Symfony\Component\ObjectMapper\TransformCallableInterface;
#[Exclude]
final class CustomerTransformer implements TransformCallableInterface
{
public function __construct(private readonly string $field = 'name') {}
public function __invoke(mixed $value, object $source, ?object $target): mixed
{
// $value = source field value, $source = full source object
$dto = new CustomerDto();
$dto->name = $value[$this->field] ?? '';
return $dto;
}
}
Use #[Exclude] so Symfony's container doesn't try to autowire transformer constructor args.
Input DTOs per Operation
When the write model differs from the read model, give individual operations their
own input: DTO and processor (the resource class stays the read model):
#[ApiResource(
operations: [
new Post(input: CreateOrderInput::class, processor: OrderCreateProcessor::class),
new Patch(input: UpdateOrderInput::class, processor: OrderUpdateProcessor::class),
]
)]
class Order { /* read model */ }
Nested Sub-Resources with uriVariables
For resources nested under parents (e.g., /accounts/{accountId}/mailboxes/{mailboxId}/messages):
#[ApiResource(
operations: [
new Get(
uriTemplate: '/accounts/{accountId}/mailboxes/{mailboxId}/messages/{id}',
uriVariables: ['accountId', 'mailboxId', 'id'],
stateOptions: new Options(
handleLinks: MessageLinkHandler::class,
documentClass: Message::class,
)
),
new GetCollection(
uriTemplate: '/accounts/{accountId}/mailboxes/{mailboxId}/messages',
uriVariables: ['accountId', 'mailboxId'],
stateOptions: new Options(
handleLinks: MessageLinkHandler::class,
documentClass: Message::class,
),
itemUriTemplate: '/accounts/{accountId}/mailboxes/{mailboxId}/messages/{id}',
),
]
)]
The handleLinks class validates parent ownership and applies security filters. See the securing-collections skill for implementation details.
Hidden IRI fields for URI generation:
#[ApiProperty(readable: false, writable: false)]
#[Map(source: 'account', transform: new DocumentIdTransformer())]
public string $accountId;
Custom Named Operations
Add non-CRUD actions on the same resource:
new Put(
uriTemplate: '/orders/{id}/cancel',
input: CancelOrderInput::class,
processor: OrderCancelProcessor::class,
name: '_api_order_cancel',
),
new Get(
write: true, // triggers processor on GET
uriTemplate: '/orders/{id}/download',
processor: OrderDownloadProcessor::class,
name: '_api_order_download',
),
Use write: true on Get operations that need a processor (e.g., file downloads):
a Get's write phase is off by default, so the processor never fires until you flip
the flag. See operations → Read & write phases for the read/write lifecycle.
Backed Enums
As a property
A backed enum property serializes to its ->value automatically:
#[ApiResource]
class Person
{
public GenderType $genderType; // backed enum → {"genderType": "female"}
}
As a resource
Expose a backed enum as a read-only resource so clients can discover allowed values:
use ApiPlatform\Metadata\ApiResource;
#[ApiResource]
enum AvailabilityStatus: string
{
case InStock = 'InStock';
case OutOfStock = 'OutOfStock';
}
GET /availability_statuses lists all cases; GET /availability_statuses/InStock
returns one.
High-Precision Numbers
For monetary/scientific values that must avoid float drift, type a property as
\BcMath\Number (PHP 8.4 native, requires ext-bcmath). It serializes as a string
to preserve precision:
class Invoice
{
public ?\BcMath\Number $total; // → "300.55"
}
Laravel (Eloquent)
The design-first split, DTOs, Object Mapper, per-operation input:, nested
uriVariables/handleLinks, custom operations and backed-enum resources all work on
Laravel. Deltas:
- Model as resource:
#[ApiResource]on a class extendingIlluminate\Database\Eloquent\Model; the Eloquent provider/processor handle CRUD with zero wiring. - Eloquent
Options(ApiPlatform\Laravel\Eloquent\State\Options) usesmodelClass:(notentityClass:/documentClass:); it carriesmodelClass+handleLinksonly — norepositoryMethod. - Object Mapper is supported: pair
stateOptions: new Options(modelClass: ProductModel::class)with#[Map(source: ProductModel::class)]on the DTO; per-property#[Map]andTransformCallableInterfaceare identical. - Properties: Eloquent models have no typed properties, so declare
#[ApiProperty]at the class level withproperty:(#[ApiProperty(property: 'title', identifier: true)]). DTO/ApiResourceclasses use property-level attributes.\BcMath\Numberand backed enums behave the same.
Checklist
When creating a new resource:
- Create the DTO class in
src/ApiResource/ - Add
#[ApiResource]with operations - Define provider for read operations (or use
stateOptions) - Define processor for write operations
- Add validation constraints to input DTOs
- For nested resources: configure
uriVariablesandhandleLinks - Add
#[Map]attributes for entity/document mapping - Add
#[ApiProperty]for OpenAPI documentation
Dateien in diesem Skill
1api-resource Skill-Fragen
Wie installiere ich den api-resource Skill in Claude Code?
Führe npx degit api-platform/skillset/skills/api-resource ~/.claude/skills/api-resource aus. Claude Code lädt Skills aus ~/.claude/skills (alle Projekte) oder .claude/skills (ein Projekt) und nutzt einen Skill, wenn seine Beschreibung zur Aufgabe passt.
Welche Clients können den api-resource Skill nutzen?
Jeder Client, der das Agent Skills-Format liest: Claude Code, Codex, Gemini CLI und claude.ai (als ZIP hochladen). Skripte im Skill laufen nur dort, wo der Client Codeausführung erlaubt.
When should I use a DTO instead of an entity as a resource?
Use a DTO when the public API shape needs to be decoupled from your database schema, especially if fields need renaming or transforming.
How do I enable the Object Mapper in API Platform?
The Object Mapper activates when you provide both stateOptions naming the entity or document class and a #[Map] attribute on the DTO.
How can I handle high-precision numbers like currency?
Type the property as \BcMath\Number to ensure it serializes as a string and avoids float drift.