Use when working with AI schema generation, agent execution, pipeline orchestration, vector storage, agent tools, or files in packages/ai-schema/, packages/ai-agent/, packages/ai-pipeline/, packages/ai-vector/, packages/ai-tools/, packages/ai-observability/
Use when editing docs/specs/, CLAUDE.md orchestration, or agent rules — keep subsystem specs aligned with code, run drift checks, and follow the anchor-issue + design-first workflow (GitHub as integration surface).
Use when working with JSON:API endpoints, resource serialization, query parsing, schema presentation, route building, or files in packages/api/, packages/routing/
Use when working with entity types, entity storage, entity queries, field definitions, field type plugins, config entities, config import/export, or files in packages/entity/, packages/entity-storage/, packages/field/, packages/config/. Covers EntityInterface, EntityBase, EntityType, EntityTypeManager, SqlEntityStorage, SqlEntityQuery, SqlSchemaHandler, EntityRepository, UnitOfWork, FieldDefinition, FieldTypeManager, FieldItemBase, ConfigFactory, ConfigManager. Also applies when touching User (packages/user/) or Node (packages/node/) entity subclasses.
Use when working with service providers, domain events, cache backends, database queries, migrations, package discovery, attribute scanning, or files in packages/foundation/, packages/cache/, packages/database-legacy/, packages/plugin/
Use when working with HTTP middleware, event middleware, job middleware, pipeline compilation, middleware discovery, or files in packages/foundation/src/Middleware/, packages/routing/, packages/user/src/Middleware/, packages/access/src/Middleware/, public/index.php
First fixture skill used by the client-transformer contract tests.
Second fixture skill — single-paragraph body for size assertions.