Implement composable filters and router-service-repository structure for paginated Drizzle list APIs.
npx skills add https://github.com/johinsDev/loyalty-app --skill api-filters
Please help me install this Agent Skill. Skill: api-filters Source: https://github.com/johinsDev/loyalty-app/tree/main/.claude/skills/api-filters Command: npx skills add https://github.com/johinsDev/loyalty-app --skill api-filters
It solves the problem of messy, duplicated list-query logic by providing a consistent three-layer feature structure and a reusable filtering mechanism for paginated endpoints.
{ rows, total } where the total matches the active filters, using the same filter set for both queries.features/<name>/ folder shape, then add new list filters without touching unrelated router/service code.Tell the AI to scaffold a new @loyalty/api feature folder using the filters pattern so a list endpoint supports filterable search and returns paginated rows plus a filter-aware total count.
High-intent search queries and answers about installing and using this skill.
Structure tRPC list endpoints using a router-service-repository pattern where composable filter classes handle Drizzle query chaining. This separates tRPC wiring, business rules, and database access into clean layers for easier testing and evolution.
The repository pattern for tRPC and Drizzle pagination isolates database access in repository classes using Drizzle's $dynamic() method. This allows services to instantiate repositories per request, ensuring safe context-bound database usage while maintaining clean separation.
Keep filtered total counts consistent by applying the same composable filter set to both row queries and total-count queries. This ensures the returned total matches the active filters, providing accurate pagination metadata for the displayed results.
Yes, you can refactor inline Drizzle routers into a features/<name>/ folder structure containing router, service, and repository files. This allows you to add new list filters without touching unrelated router or service code, improving maintainability.
Yes, the composable filter approach works without external dependencies by implementing a shared Filters base class with allowed filter keys and per-feature filter subclasses. It relies solely on tRPC and Drizzle ORM already present in your stack.
Use dynamic query building for API filters in Drizzle when adding or refactoring list endpoints with optional filters. Drizzle's $dynamic() method enables deterministic query chaining, allowing composable filter classes to conditionally apply where clauses.
jupremDefine NeonDB schemas with Drizzle ORM relations and drizzle-kit migrations.
caobingshengRetrieve paginated multi-table data with vk.baseDao.getTableData.
PlazaCCImplement JWT and cookie authentication for tRPC across HTTP, WebSocket, and SSE transports.
helixboxConstruct schema-compliant PubFi DSL requests for OpenSearch and Postgres.