Version 1.1.2

* feat: add opt-in default_load_options parameter in CrudFactory

* docs: add Relationship loading in CRUD

.github/workflows/docs.yml

@@ -1,12 +1,14 @@
 name: Documentation
+
 on:
-  push:
+  release:
-    branches:
+    types: [published]
-      - main
+
 permissions:
   contents: read
   pages: write
   id-token: write
+
 jobs:
   deploy:
     environment:

docs/module/cli.md

@@ -1,6 +1,6 @@
 # CLI
 
-Typer-based command-line interface for managing your FastAPI application, with built-in fixture loading.
+Typer-based command-line interface for managing your FastAPI application, with built-in fixture commands integration.
 
 ## Installation
 
@@ -16,7 +16,7 @@ Typer-based command-line interface for managing your FastAPI application, with b
 
 ## Overview
 
-The `cli` module provides a `manager` entry point built with [Typer](https://typer.tiangolo.com/). It auto-discovers fixture commands when a [`FixtureRegistry`](../reference/fixtures.md#fastapi_toolsets.fixtures.registry.FixtureRegistry) and a database context are configured.
+The `cli` module provides a `manager` entry point built with [Typer](https://typer.tiangolo.com/). It allow custom commands to be added in addition of the fixture commands when a [`FixtureRegistry`](../reference/fixtures.md#fastapi_toolsets.fixtures.registry.FixtureRegistry) and a database context are configured.
 
 ## Configuration
 
@@ -24,24 +24,48 @@ Configure the CLI in your `pyproject.toml`:
 
 ```toml
 [tool.fastapi-toolsets]
-cli = "myapp.cli:cli"           # optional: your custom Typer app
+cli = "myapp.cli:cli"                   # Custom Typer app
-fixtures = "myapp.fixtures:registry"  # FixtureRegistry instance
+fixtures = "myapp.fixtures:registry"    # FixtureRegistry instance
-db_context = "myapp.db:db_context"    # async context manager for sessions
+db_context = "myapp.db:db_context"      # Async context manager for sessions
 ```
 
-All fields are optional. Without configuration, the `manager` command still works but only includes the built-in commands.
+All fields are optional. Without configuration, the `manager` command still works but no command are available.
 
 ## Usage
 
 ```bash
-# List available commands
+# Manager commands
 manager --help
 
-# Load fixtures for a specific context
+ Usage: manager [OPTIONS] COMMAND [ARGS]...
-manager fixtures load --context testing
 
-# Load all fixtures (no context filter)
+ FastAPI utilities CLI.
-manager fixtures load
+
+╭─ Options ────────────────────────────────────────────────────────────────────────╮
+│ --install-completion          Install completion for the current shell.          │
+│ --show-completion             Show completion for the current shell, to copy it  │
+│                               or customize the installation.                     │
+│ --help                        Show this message and exit.                        │
+╰──────────────────────────────────────────────────────────────────────────────────╯
+╭─ Commands ───────────────────────────────────────────────────────────────────────╮
+│ check-db                                                                         │
+│ fixtures  Manage database fixtures.                                              │
+╰──────────────────────────────────────────────────────────────────────────────────╯
+
+# Fixtures commands
+manager fixtures --help
+
+ Usage: manager fixtures [OPTIONS] COMMAND [ARGS]...
+
+ Manage database fixtures.
+
+╭─ Options ────────────────────────────────────────────────────────────────────────╮
+│ --help          Show this message and exit.                                      │
+╰──────────────────────────────────────────────────────────────────────────────────╯
+╭─ Commands ───────────────────────────────────────────────────────────────────────╮
+│ list  List all registered fixtures.                                              │
+│ load  Load fixtures into the database.                                           │
+╰──────────────────────────────────────────────────────────────────────────────────╯
 ```
 
 ## Custom CLI
@@ -64,14 +88,6 @@ def hello():
 cli = "myapp.cli:cli"
 ```
 
-## Entry point
-
-The `manager` script is registered automatically when the package is installed:
-
-```bash
-manager --help
-```
-
 ---
 
 [:material-api: API Reference](../reference/cli.md)

docs/module/crud.md

@@ -2,6 +2,9 @@
 
 Generic async CRUD operations for SQLAlchemy models with search, pagination, and many-to-many support.
 
+!!! info
+    This module has been coded and tested to be compatible with PostgreSQL only.
+
 ## Overview
 
 The `crud` module provides [`AsyncCrud`](../reference/crud.md#fastapi_toolsets.crud.factory.AsyncCrud), an abstract base class with a full suite of async database operations, and [`CrudFactory`](../reference/crud.md#fastapi_toolsets.crud.factory.CrudFactory), a convenience function to instantiate it for a given model.
@@ -12,10 +15,7 @@ The `crud` module provides [`AsyncCrud`](../reference/crud.md#fastapi_toolsets.c
 from fastapi_toolsets.crud import CrudFactory
 from myapp.models import User
 
-UserCrud = CrudFactory(
+UserCrud = CrudFactory(model=User)
-    User,
-    searchable_fields=[User.username, User.email],
-)
 ```
 
 [`CrudFactory`](../reference/crud.md#fastapi_toolsets.crud.factory.CrudFactory) dynamically creates a class named `AsyncUserCrud` with `User` as its model.
@@ -24,42 +24,146 @@ UserCrud = CrudFactory(
 
 ```python
 # Create
-user = await UserCrud.create(session, obj=UserCreateSchema(username="alice"))
+user = await UserCrud.create(session=session, obj=UserCreateSchema(username="alice"))
 
 # Get one (raises NotFoundError if not found)
-user = await UserCrud.get(session, filters=[User.id == user_id])
+user = await UserCrud.get(session=session, filters=[User.id == user_id])
 
 # Get first or None
-user = await UserCrud.first(session, filters=[User.email == email])
+user = await UserCrud.first(session=session, filters=[User.email == email])
 
 # Get multiple
-users = await UserCrud.get_multi(session, filters=[User.is_active == True])
+users = await UserCrud.get_multi(session=session, filters=[User.is_active == True])
 
 # Update
-user = await UserCrud.update(session, obj=UserUpdateSchema(username="bob"), filters=[User.id == user_id])
+user = await UserCrud.update(session=session, obj=UserUpdateSchema(username="bob"), filters=[User.id == user_id])
 
 # Delete
-await UserCrud.delete(session, filters=[User.id == user_id])
+await UserCrud.delete(session=session, filters=[User.id == user_id])
 
 # Count / exists
-count = await UserCrud.count(session, filters=[User.is_active == True])
+count = await UserCrud.count(session=session, filters=[User.is_active == True])
-exists = await UserCrud.exists(session, filters=[User.email == email])
+exists = await UserCrud.exists(session=session, filters=[User.email == email])
 ```
 
 ## Pagination
 
+!!! info "Added in `v1.1` (only offset_pagination via `paginate` if `<v1.1`)"
+
+Two pagination strategies are available. Both return a [`PaginatedResponse`](../reference/schemas.md#fastapi_toolsets.schemas.PaginatedResponse) but differ in how they navigate through results.
+
+| | `offset_paginate` | `cursor_paginate` |
+|---|---|---|
+| Total count | Yes | No |
+| Jump to arbitrary page | Yes | No |
+| Performance on deep pages | Degrades | Constant |
+| Stable under concurrent inserts | No | Yes |
+| Search compatible | Yes | Yes |
+| Use case | Admin panels, numbered pagination | Feeds, APIs, infinite scroll |
+
+### Offset pagination
+
 ```python
-result = await UserCrud.paginate(
+@router.get(
-    session,
+    "",
-    filters=[User.is_active == True],
+    response_model=PaginatedResponse[User],
-    order_by=[User.created_at.desc()],
-    page=1,
-    items_per_page=20,
-    search="alice",
-    search_fields=[User.username, User.email],
 )
-# result.data: list of users
+async def get_users(
-# result.pagination: Pagination(total_count, items_per_page, page, has_more)
+    session: SessionDep,
+    items_per_page: int = 50,
+    page: int = 1,
+):
+    return await crud.UserCrud.offset_paginate(
+        session=session,
+        items_per_page=items_per_page,
+        page=page,
+    )
+```
+
+The [`offset_paginate`](../reference/crud.md#fastapi_toolsets.crud.factory.AsyncCrud.offset_paginate) method returns a [`PaginatedResponse`](../reference/schemas.md#fastapi_toolsets.schemas.PaginatedResponse) whose `pagination` field is an [`OffsetPagination`](../reference/schemas.md#fastapi_toolsets.schemas.OffsetPagination) object:
+
+```json
+{
+  "status": "SUCCESS",
+  "data": ["..."],
+  "pagination": {
+    "total_count": 100,
+    "page": 1,
+    "items_per_page": 20,
+    "has_more": true
+  }
+}
+```
+
+!!! warning "Deprecated: `paginate`"
+    The `paginate` function is a backward-compatible alias for `offset_paginate`. This function is **deprecated** and will be removed in **v2.0**.
+
+### Cursor pagination
+
+```python
+@router.get(
+    "",
+    response_model=PaginatedResponse[UserRead],
+)
+async def list_users(
+    session: SessionDep,
+    cursor: str | None = None,
+    items_per_page: int = 20,
+):
+    return await UserCrud.cursor_paginate(
+        session=session,
+        cursor=cursor,
+        items_per_page=items_per_page,
+    )
+```
+
+The [`cursor_paginate`](../reference/crud.md#fastapi_toolsets.crud.factory.AsyncCrud.cursor_paginate) method returns a [`PaginatedResponse`](../reference/schemas.md#fastapi_toolsets.schemas.PaginatedResponse) whose `pagination` field is a [`CursorPagination`](../reference/schemas.md#fastapi_toolsets.schemas.CursorPagination) object:
+
+```json
+{
+  "status": "SUCCESS",
+  "data": ["..."],
+  "pagination": {
+    "next_cursor": "eyJ2YWx1ZSI6ICIzZjQ3YWM2OS0uLi4ifQ==",
+    "prev_cursor": null,
+    "items_per_page": 20,
+    "has_more": true
+  }
+}
+```
+
+Pass `next_cursor` as the `cursor` query parameter on the next request to advance to the next page. `prev_cursor` is set on pages 2+ and points back to the first item of the current page. Both are `null` when there is no adjacent page.
+
+#### Choosing a cursor column
+
+The cursor column is set once on [`CrudFactory`](../reference/crud.md#fastapi_toolsets.crud.factory.CrudFactory) via the `cursor_column` parameter. It must be monotonically ordered for stable results:
+
+- Auto-increment integer PKs
+- UUID v7 PKs
+- Timestamps
+
+!!! warning
+    Random UUID v4 PKs are **not** suitable as cursor columns because their ordering is non-deterministic.
+
+!!! note
+    `cursor_column` is required. Calling [`cursor_paginate`](../reference/crud.md#fastapi_toolsets.crud.factory.AsyncCrud.cursor_paginate) on a CRUD class that has no `cursor_column` configured raises a `ValueError`.
+
+The cursor value is base64-encoded when returned to the client and decoded back to the correct Python type on the next request. The following SQLAlchemy column types are supported:
+
+| SQLAlchemy type | Python type |
+|---|---|
+| `Integer`, `BigInteger`, `SmallInteger` | `int` |
+| `Uuid` | `uuid.UUID` |
+| `DateTime` | `datetime.datetime` |
+| `Date` | `datetime.date` |
+| `Float`, `Numeric` | `decimal.Decimal` |
+
+```python
+# Paginate by the primary key
+PostCrud = CrudFactory(model=Post, cursor_column=Post.id)
+
+# Paginate by a timestamp column instead
+PostCrud = CrudFactory(model=Post, cursor_column=Post.created_at)
 ```
 
 ## Search
@@ -68,7 +172,7 @@ Declare searchable fields on the CRUD class. Relationship traversal is supported
 
 ```python
 PostCrud = CrudFactory(
-    Post,
+    model=Post,
     searchable_fields=[
         Post.title,
         Post.content,
@@ -77,18 +181,92 @@ PostCrud = CrudFactory(
 )
 ```
 
+This allows searching with both [`offset_paginate`](../reference/crud.md#fastapi_toolsets.crud.factory.AsyncCrud.offset_paginate) and [`cursor_paginate`](../reference/crud.md#fastapi_toolsets.crud.factory.AsyncCrud.cursor_paginate):
+
+```python
+@router.get(
+    "",
+    response_model=PaginatedResponse[User],
+)
+async def get_users(
+    session: SessionDep,
+    items_per_page: int = 50,
+    page: int = 1,
+    search: str | None = None,
+):
+    return await crud.UserCrud.offset_paginate(
+        session=session,
+        items_per_page=items_per_page,
+        page=page,
+        search=search,
+    )
+```
+
+```python
+@router.get(
+    "",
+    response_model=PaginatedResponse[User],
+)
+async def get_users(
+    session: SessionDep,
+    cursor: str | None = None,
+    items_per_page: int = 50,
+    search: str | None = None,
+):
+    return await crud.UserCrud.cursor_paginate(
+        session=session,
+        items_per_page=items_per_page,
+        cursor=cursor,
+        search=search,
+    )
+```
+
+## Relationship loading
+
+!!! info "Added in `v1.1`"
+
+By default, SQLAlchemy relationships are not loaded unless explicitly requested. Instead of using `lazy="selectin"` on model definitions (which is implicit and applies globally), define a `default_load_options` on the CRUD class to control loading strategy explicitly.
+
+!!! warning
+    Avoid using `lazy="selectin"` on model relationships. It fires silently on every query, cannot be disabled per-call, and can cause unexpected cascading loads through deep relationship chains. Use `default_load_options` instead.
+
+```python
+from sqlalchemy.orm import selectinload
+
+ArticleCrud = CrudFactory(
+    model=Article,
+    default_load_options=[
+        selectinload(Article.category),
+        selectinload(Article.tags),
+    ],
+)
+```
+
+`default_load_options` applies automatically to all read operations (`get`, `first`, `get_multi`, `offset_paginate`, `cursor_paginate`). When `load_options` is passed at call-site, it **fully replaces** `default_load_options` for that query — giving you precise per-call control:
+
+```python
+# Only loads category, tags are not loaded
+article = await ArticleCrud.get(
+    session=session,
+    filters=[Article.id == article_id],
+    load_options=[selectinload(Article.category)],
+)
+
+# Loads nothing — useful for write-then-refresh flows or lightweight checks
+articles = await ArticleCrud.get_multi(session=session, load_options=[])
+```
+
 ## Many-to-many relationships
 
 Use `m2m_fields` to map schema fields containing lists of IDs to SQLAlchemy relationships. The CRUD class resolves and validates all IDs before persisting:
 
 ```python
 PostCrud = CrudFactory(
-    Post,
+    model=Post,
     m2m_fields={"tag_ids": Post.tags},
 )
 
-# schema: PostCreateSchema(title="Hello", tag_ids=[1, 2, 3])
+post = await PostCrud.create(session=session, obj=PostCreateSchema(title="Hello", tag_ids=[1, 2, 3]))
-post = await PostCrud.create(session, obj=PostCreateSchema(...))
 ```
 
 ## Upsert
@@ -97,20 +275,49 @@ Atomic `INSERT ... ON CONFLICT DO UPDATE` using PostgreSQL:
 
 ```python
 await UserCrud.upsert(
-    session,
+    session=session,
     obj=UserCreateSchema(email="alice@example.com", username="alice"),
     index_elements=[User.email],
     set_={"username"},
 )
 ```
 
-## `as_response`
+## `schema` — typed response serialization
 
-Pass `as_response=True` to any write operation to get a [`Response[ModelType]`](../reference/schemas.md#fastapi_toolsets.schemas.Response) back directly:
+!!! info "Added in `v1.1`"
+
+Pass a Pydantic schema class to `create`, `get`, `update`, or `offset_paginate` to serialize the result directly into that schema and wrap it in a [`Response[schema]`](../reference/schemas.md#fastapi_toolsets.schemas.Response) or [`PaginatedResponse[schema]`](../reference/schemas.md#fastapi_toolsets.schemas.PaginatedResponse):
 
 ```python
-response = await UserCrud.create(session, obj=schema, as_response=True)
+class UserRead(PydanticBase):
-# response: Response[User]
+    id: UUID
+    username: str
+
+@router.get(
+    "/{uuid}",
+    responses=generate_error_responses(NotFoundError),
+)
+async def get_user(session: SessionDep, uuid: UUID) -> Response[UserRead]:
+    return await crud.UserCrud.get(
+        session=session,
+        filters=[User.id == uuid],
+        schema=UserRead,
+    )
+
+@router.get("")
+async def list_users(session: SessionDep, page: int = 1) -> PaginatedResponse[UserRead]:
+    return await crud.UserCrud.offset_paginate(
+        session=session,
+        page=page,
+        schema=UserRead,
+    )
 ```
 
+The schema must have `from_attributes=True` (or inherit from [`PydanticBase`](../reference/schemas.md#fastapi_toolsets.schemas.PydanticBase)) so it can be built from SQLAlchemy model instances.
+
+!!! warning "Deprecated: `as_response`"
+    The `as_response=True` parameter is **deprecated** and will be removed in **v2.0**. Replace it with `schema=YourSchema`.
+
+---
+
 [:material-api: API Reference](../reference/crud.md)

docs/module/db.md

@@ -2,9 +2,12 @@
 
 SQLAlchemy async session management with transactions, table locking, and row-change polling.
 
+!!! info
+    This module has been coded and tested to be compatible with PostgreSQL only.
+
 ## Overview
 
-The `db` module provides helpers to create FastAPI dependencies and context managers for `AsyncSession`, along with utilities for nested transactions, PostgreSQL advisory locks, and polling for row changes.
+The `db` module provides helpers to create FastAPI dependencies and context managers for `AsyncSession`, along with utilities for nested transactions, table lock and polling for row changes.
 
 ## Session dependency
 
@@ -14,10 +17,10 @@ Use [`create_db_dependency`](../reference/db.md#fastapi_toolsets.db.create_db_de
 from sqlalchemy.ext.asyncio import create_async_engine, async_sessionmaker
 from fastapi_toolsets.db import create_db_dependency
 
-engine = create_async_engine("postgresql+asyncpg://...")
+engine = create_async_engine(url="postgresql+asyncpg://...", future=True)
-session_maker = async_sessionmaker(engine)
+session_maker = async_sessionmaker(bind=engine, expire_on_commit=False)
 
-get_db = create_db_dependency(session_maker)
+get_db = create_db_dependency(session_maker=session_maker)
 
 @router.get("/users")
 async def list_users(session: AsyncSession = Depends(get_db)):
@@ -31,11 +34,11 @@ Use [`create_db_context`](../reference/db.md#fastapi_toolsets.db.create_db_conte
 ```python
 from fastapi_toolsets.db import create_db_context
 
-db_context = create_db_context(session_maker)
+db_context = create_db_context(session_maker=session_maker)
 
 async def seed():
     async with db_context() as session:
-        session.add(User(name="admin"))
+        ...
 ```
 
 ## Nested transactions
@@ -45,11 +48,11 @@ async def seed():
 ```python
 from fastapi_toolsets.db import get_transaction
 
-async def create_user_with_role(session):
+async def create_user_with_role(session=session):
-    async with get_transaction(session):
+    async with get_transaction(session=session):
-        session.add(role)
+        ...
-        async with get_transaction(session):  # uses savepoint
+        async with get_transaction(session=session):  # uses savepoint
-            session.add(user)
+            ...
 ```
 
 ## Table locking
@@ -59,7 +62,7 @@ async def create_user_with_role(session):
 ```python
 from fastapi_toolsets.db import lock_tables
 
-async with lock_tables(session, tables=[User], mode="EXCLUSIVE"):
+async with lock_tables(session=session, tables=[User], mode="EXCLUSIVE"):
     # No other transaction can modify User until this block exits
     ...
 ```
@@ -75,7 +78,7 @@ from fastapi_toolsets.db import wait_for_row_change
 
 # Wait up to 30s for order.status to change
 await wait_for_row_change(
-    session,
+    session=session,
     model=Order,
     pk_value=order_id,
     columns=[Order.status],

docs/module/dependencies.md

@@ -13,17 +13,17 @@ The `dependencies` module provides two factory functions that create FastAPI dep
 ```python
 from fastapi_toolsets.dependencies import PathDependency
 
-UserDep = PathDependency(User, User.id, session_dep=get_db)
+UserDep = PathDependency(model=User, field=User.id, session_dep=get_db)
 
 @router.get("/users/{user_id}")
 async def get_user(user: User = UserDep):
     return user
 ```
 
-The parameter name is inferred from the field (`user_id` for `User.id`). You can override it:
+By default the parameter name is inferred from the field (`user_id` for `User.id`). You can override it:
 
 ```python
-UserDep = PathDependency(User, User.id, session_dep=get_db, param_name="id")
+UserDep = PathDependency(model=User, field=User.id, session_dep=get_db, param_name="id")
 
 @router.get("/users/{id}")
 async def get_user(user: User = UserDep):
@@ -37,7 +37,7 @@ async def get_user(user: User = UserDep):
 ```python
 from fastapi_toolsets.dependencies import BodyDependency
 
-RoleDep = BodyDependency(Role, Role.id, session_dep=get_db, body_field="role_id")
+RoleDep = BodyDependency(model=Role, field=Role.id, session_dep=get_db, body_field="role_id")
 
 @router.post("/users")
 async def create_user(body: UserCreateSchema, role: Role = RoleDep):
@@ -45,4 +45,6 @@ async def create_user(body: UserCreateSchema, role: Role = RoleDep):
     ...
 ```
 
+---
+
 [:material-api: API Reference](../reference/dependencies.md)

docs/module/exceptions.md

@@ -4,7 +4,7 @@ Structured API exceptions with consistent error responses and automatic OpenAPI
 
 ## Overview
 
-The `exceptions` module provides a set of pre-built HTTP exceptions and a FastAPI exception handler that formats all errors — including validation errors — into a uniform [`ErrorResponse`](../reference/schemas.md#fastapi_toolsets.schemas.ErrorResponse) shape.
+The `exceptions` module provides a set of pre-built HTTP exceptions and a FastAPI exception handler that formats all errors — including validation errors — into a uniform [`ErrorResponse`](../reference/schemas.md#fastapi_toolsets.schemas.ErrorResponse).
 
 ## Setup
 
@@ -15,7 +15,7 @@ from fastapi import FastAPI
 from fastapi_toolsets.exceptions import init_exceptions_handlers
 
 app = FastAPI()
-init_exceptions_handlers(app)
+init_exceptions_handlers(app=app)
 ```
 
 This registers handlers for:
@@ -36,11 +36,11 @@ This registers handlers for:
 | [`NoSearchableFieldsError`](../reference/exceptions.md#fastapi_toolsets.exceptions.exceptions.NoSearchableFieldsError) | 400 | No searchable fields |
 
 ```python
-from fastapi_toolsets.exceptions import NotFoundError, ForbiddenError
+from fastapi_toolsets.exceptions import NotFoundError
 
 @router.get("/users/{id}")
 async def get_user(id: int, session: AsyncSession = Depends(get_db)):
-    user = await UserCrud.first(session, filters=[User.id == id])
+    user = await UserCrud.first(session=session, filters=[User.id == id])
     if not user:
         raise NotFoundError
     return user
@@ -77,6 +77,9 @@ from fastapi_toolsets.exceptions import generate_error_responses, NotFoundError,
 async def get_user(...): ...
 ```
 
+!!! info
+    The pydantic validation error is automatically added by FastAPI.
+
 ---
 
 [:material-api: API Reference](../reference/exceptions.md)

docs/module/fixtures.md

@@ -32,22 +32,22 @@ Dependencies declared via `depends_on` are resolved topologically — `roles` wi
 
 ## Loading fixtures
 
-### By context
+By context with [`load_fixtures_by_context`](../reference/fixtures.md#fastapi_toolsets.fixtures.utils.load_fixtures_by_context):
 
 ```python
 from fastapi_toolsets.fixtures import load_fixtures_by_context
 
 async with db_context() as session:
-    await load_fixtures_by_context(session, registry=fixtures, context=Context.TESTING)
+    await load_fixtures_by_context(session=session, registry=fixtures, context=Context.TESTING)
 ```
 
-### Directly
+Directly with [`load_fixtures`](../reference/fixtures.md#fastapi_toolsets.fixtures.utils.load_fixtures):
 
 ```python
 from fastapi_toolsets.fixtures import load_fixtures
 
 async with db_context() as session:
-    await load_fixtures(session, registry=fixtures)
+    await load_fixtures(session=session, registry=fixtures)
 ```
 
 ## Contexts
@@ -60,7 +60,7 @@ async with db_context() as session:
 | `Context.TESTING` | Data only loaded during tests |
 | `Context.PRODUCTION` | Data only loaded in production |
 
-A fixture with no `contexts` argument is loaded in all contexts.
+A fixture with no `contexts` defined takes `Context.BASE` by default.
 
 ## Load strategies
 
@@ -72,9 +72,21 @@ A fixture with no `contexts` argument is loaded in all contexts.
 | `LoadStrategy.UPSERT` | Insert or update on conflict |
 | `LoadStrategy.SKIP` | Skip rows that already exist |
 
+## Merging registries
+
+Split fixtures definitions across modules and merge them:
+
+```python
+from myapp.fixtures.dev import dev_fixtures
+from myapp.fixtures.prod import prod_fixtures
+
+fixtures = fixturesRegistry()
+fixtures.include_registry(registry=dev_fixtures)
+fixtures.include_registry(registry=prod_fixtures)
+
 ## Pytest integration
 
-Use [`register_fixtures`](../reference/pytest.md#fastapi_toolsets.pytest.plugin.register_fixtures) to expose each fixture in your registry as an injectable pytest fixture named `fixture_{name}`:
+Use [`register_fixtures`](../reference/pytest.md#fastapi_toolsets.pytest.plugin.register_fixtures) to expose each fixture in your registry as an injectable pytest fixture named `fixture_{name}` by default:
 
 ```python
 # conftest.py
@@ -95,10 +107,8 @@ register_fixtures(registry=registry, namespace=globals())
 
 ```python
 # test_users.py
-async def test_user_can_login(fixture_users, fixture_roles, client):
+async def test_user_can_login(fixture_users: list[User], fixture_roles: list[Role]):
-    # fixture_roles is loaded first (dependency), then fixture_users
+    ...
-    response = await client.post("/auth/login", json={"username": "alice"})
-    assert response.status_code == 200
 ```
 
 

docs/module/logger.md

@@ -34,4 +34,6 @@ When called without arguments, [`get_logger`](../reference/logger.md#fastapi_too
 logger = get_logger()
 ```
 
+---
+
 [:material-api: API Reference](../reference/logger.md)

docs/module/metrics.md

@@ -27,7 +27,7 @@ from fastapi_toolsets.metrics import MetricsRegistry, init_metrics
 app = FastAPI()
 metrics = MetricsRegistry()
 
-init_metrics(app, registry=metrics)
+init_metrics(app=app, registry=metrics)
 ```
 
 This mounts the `/metrics` endpoint that Prometheus can scrape.
@@ -70,8 +70,8 @@ from myapp.metrics.http import http_metrics
 from myapp.metrics.db import db_metrics
 
 metrics = MetricsRegistry()
-metrics.include_registry(http_metrics)
+metrics.include_registry(registry=http_metrics)
-metrics.include_registry(db_metrics)
+metrics.include_registry(registry=db_metrics)
 ```
 
 ## Multi-process mode
@@ -81,10 +81,6 @@ Multi-process support is enabled automatically when the `PROMETHEUS_MULTIPROC_DI
 !!! warning "Environment variable name"
     The correct variable is `PROMETHEUS_MULTIPROC_DIR` (not `PROMETHEUS_MULTIPROCESS_DIR`).
 
-```bash
-export PROMETHEUS_MULTIPROC_DIR=/tmp/prometheus
-```
-
 ---
 
 [:material-api: API Reference](../reference/metrics.md)

docs/module/pytest.md

@@ -26,8 +26,15 @@ Use [`create_async_client`](../reference/pytest.md#fastapi_toolsets.pytest.utils
 from fastapi_toolsets.pytest import create_async_client
 
 @pytest.fixture
-async def client(app):
+async def http_client(db_session):
-    async with create_async_client(app=app) as c:
+    async def _override_get_db():
+        yield db_session
+
+    async with create_async_client(
+        app=app,
+        base_url="http://127.0.0.1/api/v1",
+        dependency_overrides={get_db: _override_get_db},
+    ) as c:
         yield c
 ```
 
@@ -36,36 +43,34 @@ async def client(app):
 Use [`create_db_session`](../reference/pytest.md#fastapi_toolsets.pytest.utils.create_db_session) to create an isolated `AsyncSession` for a test:
 
 ```python
-from fastapi_toolsets.pytest import create_db_session
+from fastapi_toolsets.pytest import create_db_session, create_worker_database
 
-@pytest.fixture
-async def db_session():
-    async with create_db_session(database_url=DATABASE_URL, base=Base, cleanup=True) as session:
-        yield session
-```
-
-## Parallel testing with pytest-xdist
-
-When running tests in parallel, each worker needs its own database. Use these helpers to create and identify worker databases:
-
-```python
-from fastapi_toolsets.pytest import create_worker_database, create_db_session
-
-# In conftest.py session-scoped fixture
 @pytest.fixture(scope="session")
 async def worker_db_url():
-    async with create_worker_database(database_url=DATABASE_URL) as url:
+    async with create_worker_database(
+        database_url=str(settings.SQLALCHEMY_DATABASE_URI)
+    ) as url:
         yield url
 
+
 @pytest.fixture
 async def db_session(worker_db_url):
-    async with create_db_session(database_url=worker_db_url, base=Base, cleanup=True) as session:
+    async with create_db_session(
+        database_url=worker_db_url, base=Base, cleanup=True
+    ) as session:
         yield session
 ```
 
+!!! info
+    In this example, the database is reset between each test using the argument `cleanup=True`.
+
+## Parallel testing with pytest-xdist
+
+The examples above are already compatible with parallel test execution with `pytest-xdist`.
+
 ## Cleaning up tables
 
-[`cleanup_tables`](../reference/pytest.md#fastapi_toolsets.pytest.utils.cleanup_tables) truncates all tables between tests for fast isolation:
+If you want to manually clean up a database you can use [`cleanup_tables`](../reference/pytest.md#fastapi_toolsets.pytest.utils.cleanup_tables), this will truncates all tables between tests for fast isolation:
 
 ```python
 from fastapi_toolsets.pytest import cleanup_tables

docs/module/schemas.md

@@ -8,7 +8,7 @@ The `schemas` module provides generic response wrappers that enforce a uniform r
 
 ## Response models
 
-### `Response[T]`
+### [`Response[T]`](../reference/schemas.md#fastapi_toolsets.schemas.Response)
 
 The most common wrapper for a single resource response.
 
@@ -20,7 +20,7 @@ async def get_user(user: User = UserDep) -> Response[UserSchema]:
     return Response(data=user, message="User retrieved")
 ```
 
-### `PaginatedResponse[T]`
+### [`PaginatedResponse[T]`](../reference/schemas.md#fastapi_toolsets.schemas.PaginatedResponse)
 
 Wraps a list of items with pagination metadata.
 
@@ -40,15 +40,10 @@ async def list_users() -> PaginatedResponse[UserSchema]:
     )
 ```
 
-### `ErrorResponse`
+### [`ErrorResponse`](../reference/schemas.md#fastapi_toolsets.schemas.ErrorResponse)
 
-Returned automatically by the exceptions handler. Can also be used as a response model for OpenAPI docs.
+Returned automatically by the exceptions handler.
 
-```python
+---
-from fastapi_toolsets.schemas import ErrorResponse
-
-@router.delete("/users/{id}", responses={404: {"model": ErrorResponse}})
-async def delete_user(...): ...
-```
 
 [:material-api: API Reference](../reference/schemas.md)

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "fastapi-toolsets"
-version = "0.10.0"
+version = "1.1.2"
 description = "Reusable tools for FastAPI: async CRUD, fixtures, CLI, and standardized responses for SQLAlchemy + PostgreSQL"
 readme = "README.md"
 license = "MIT"

src/fastapi_toolsets/__init__.py

@@ -21,4 +21,4 @@ Example usage:
         return Response(data={"user": user.username}, message="Success")
 """
 
-__version__ = "0.10.0"
+__version__ = "1.1.2"

src/fastapi_toolsets/cli/config.py

@@ -1,5 +1,7 @@
 """CLI configuration and dynamic imports."""
 
+from __future__ import annotations
+
 import importlib
 import sys
 from typing import TYPE_CHECKING, Any, Literal, overload

src/fastapi_toolsets/crud/factory.py

@@ -2,28 +2,46 @@
 
 from __future__ import annotations
 
+import base64
+import json
+import uuid as uuid_module
+import warnings
 from collections.abc import Mapping, Sequence
+from datetime import date, datetime
+from decimal import Decimal
 from typing import Any, ClassVar, Generic, Literal, Self, TypeVar, cast, overload
 
 from pydantic import BaseModel
-from sqlalchemy import and_, func, select
+from sqlalchemy import Date, DateTime, Float, Integer, Numeric, Uuid, and_, func, select
 from sqlalchemy import delete as sql_delete
 from sqlalchemy.dialects.postgresql import insert
 from sqlalchemy.exc import NoResultFound
 from sqlalchemy.ext.asyncio import AsyncSession
 from sqlalchemy.orm import DeclarativeBase, QueryableAttribute, selectinload
+from sqlalchemy.sql.base import ExecutableOption
 from sqlalchemy.sql.roles import WhereHavingRole
 
 from ..db import get_transaction
 from ..exceptions import NotFoundError
-from ..schemas import PaginatedResponse, Pagination, Response
+from ..schemas import CursorPagination, OffsetPagination, PaginatedResponse, Response
 from .search import SearchConfig, SearchFieldType, build_search_filters
 
 ModelType = TypeVar("ModelType", bound=DeclarativeBase)
+SchemaType = TypeVar("SchemaType", bound=BaseModel)
 JoinType = list[tuple[type[DeclarativeBase], Any]]
 M2MFieldType = Mapping[str, QueryableAttribute[Any]]
 
 
+def _encode_cursor(value: Any) -> str:
+    """Encode cursor column value as an base64 string."""
+    return base64.b64encode(json.dumps(str(value)).encode()).decode()
+
+
+def _decode_cursor(cursor: str) -> str:
+    """Decode cursor base64 string."""
+    return json.loads(base64.b64decode(cursor.encode()).decode())
+
+
 class AsyncCrud(Generic[ModelType]):
     """Generic async CRUD operations for SQLAlchemy models.
 
@@ -33,26 +51,17 @@ class AsyncCrud(Generic[ModelType]):
     model: ClassVar[type[DeclarativeBase]]
     searchable_fields: ClassVar[Sequence[SearchFieldType] | None] = None
     m2m_fields: ClassVar[M2MFieldType | None] = None
+    default_load_options: ClassVar[list[ExecutableOption] | None] = None
+    cursor_column: ClassVar[Any | None] = None
 
-    @overload
     @classmethod
-    async def create(  # pragma: no cover
+    def _resolve_load_options(
-        cls: type[Self],
+        cls, load_options: list[ExecutableOption] | None
-        session: AsyncSession,
+    ) -> list[ExecutableOption] | None:
-        obj: BaseModel,
+        """Return load_options if provided, else fall back to default_load_options."""
-        *,
+        if load_options is not None:
-        as_response: Literal[True],
+            return load_options
-    ) -> Response[ModelType]: ...
+        return cls.default_load_options
-
-    @overload
-    @classmethod
-    async def create(  # pragma: no cover
-        cls: type[Self],
-        session: AsyncSession,
-        obj: BaseModel,
-        *,
-        as_response: Literal[False] = ...,
-    ) -> ModelType: ...
 
     @classmethod
     async def _resolve_m2m(
@@ -110,6 +119,40 @@ class AsyncCrud(Generic[ModelType]):
             return set()
         return set(cls.m2m_fields.keys())
 
+    @overload
+    @classmethod
+    async def create(  # pragma: no cover
+        cls: type[Self],
+        session: AsyncSession,
+        obj: BaseModel,
+        *,
+        schema: type[SchemaType],
+        as_response: bool = ...,
+    ) -> Response[SchemaType]: ...
+
+    # Backward-compatible - will be removed in v2.0
+    @overload
+    @classmethod
+    async def create(  # pragma: no cover
+        cls: type[Self],
+        session: AsyncSession,
+        obj: BaseModel,
+        *,
+        as_response: Literal[True],
+        schema: None = ...,
+    ) -> Response[ModelType]: ...
+
+    @overload
+    @classmethod
+    async def create(  # pragma: no cover
+        cls: type[Self],
+        session: AsyncSession,
+        obj: BaseModel,
+        *,
+        as_response: Literal[False] = ...,
+        schema: None = ...,
+    ) -> ModelType: ...
+
     @classmethod
     async def create(
         cls: type[Self],
@@ -117,17 +160,28 @@ class AsyncCrud(Generic[ModelType]):
         obj: BaseModel,
         *,
         as_response: bool = False,
-    ) -> ModelType | Response[ModelType]:
+        schema: type[BaseModel] | None = None,
+    ) -> ModelType | Response[ModelType] | Response[Any]:
         """Create a new record in the database.
 
         Args:
             session: DB async session
             obj: Pydantic model with data to create
-            as_response: If True, wrap result in Response object
+            as_response: Deprecated. Use ``schema`` instead. Will be removed in v2.0.
+            schema: Pydantic schema to serialize the result into. When provided,
+                the result is automatically wrapped in a ``Response[schema]``.
 
         Returns:
-            Created model instance or Response wrapping it
+            Created model instance, or ``Response[schema]`` when ``schema`` is given,
+            or ``Response[ModelType]`` when ``as_response=True`` (deprecated).
         """
+        if as_response and schema is None:
+            warnings.warn(
+                "as_response is deprecated and will be removed in v2.0. "
+                "Use schema=YourSchema instead.",
+                DeprecationWarning,
+                stacklevel=2,
+            )
         async with get_transaction(session):
             m2m_exclude = cls._m2m_schema_fields()
             data = (
@@ -143,8 +197,9 @@ class AsyncCrud(Generic[ModelType]):
             session.add(db_model)
         await session.refresh(db_model)
         result = cast(ModelType, db_model)
-        if as_response:
+        if as_response or schema:
-            return Response(data=result)
+            data_out = schema.model_validate(result) if schema else result
+            return Response(data=data_out)
         return result
 
     @overload
@@ -157,8 +212,25 @@ class AsyncCrud(Generic[ModelType]):
         joins: JoinType | None = None,
         outer_join: bool = False,
         with_for_update: bool = False,
-        load_options: list[Any] | None = None,
+        load_options: list[ExecutableOption] | None = None,
+        schema: type[SchemaType],
+        as_response: bool = ...,
+    ) -> Response[SchemaType]: ...
+
+    # Backward-compatible - will be removed in v2.0
+    @overload
+    @classmethod
+    async def get(  # pragma: no cover
+        cls: type[Self],
+        session: AsyncSession,
+        filters: list[Any],
+        *,
+        joins: JoinType | None = None,
+        outer_join: bool = False,
+        with_for_update: bool = False,
+        load_options: list[ExecutableOption] | None = None,
         as_response: Literal[True],
+        schema: None = ...,
     ) -> Response[ModelType]: ...
 
     @overload
@@ -171,8 +243,9 @@ class AsyncCrud(Generic[ModelType]):
         joins: JoinType | None = None,
         outer_join: bool = False,
         with_for_update: bool = False,
-        load_options: list[Any] | None = None,
+        load_options: list[ExecutableOption] | None = None,
         as_response: Literal[False] = ...,
+        schema: None = ...,
     ) -> ModelType: ...
 
     @classmethod
@@ -184,9 +257,10 @@ class AsyncCrud(Generic[ModelType]):
         joins: JoinType | None = None,
         outer_join: bool = False,
         with_for_update: bool = False,
-        load_options: list[Any] | None = None,
+        load_options: list[ExecutableOption] | None = None,
         as_response: bool = False,
-    ) -> ModelType | Response[ModelType]:
+        schema: type[BaseModel] | None = None,
+    ) -> ModelType | Response[ModelType] | Response[Any]:
         """Get exactly one record. Raises NotFoundError if not found.
 
         Args:
@@ -196,15 +270,25 @@ class AsyncCrud(Generic[ModelType]):
             outer_join: Use LEFT OUTER JOIN instead of INNER JOIN
             with_for_update: Lock the row for update
             load_options: SQLAlchemy loader options (e.g., selectinload)
-            as_response: If True, wrap result in Response object
+            as_response: Deprecated. Use ``schema`` instead. Will be removed in v2.0.
+            schema: Pydantic schema to serialize the result into. When provided,
+                the result is automatically wrapped in a ``Response[schema]``.
 
         Returns:
-            Model instance or Response wrapping it
+            Model instance, or ``Response[schema]`` when ``schema`` is given,
+            or ``Response[ModelType]`` when ``as_response=True`` (deprecated).
 
         Raises:
             NotFoundError: If no record found
             MultipleResultsFound: If more than one record found
         """
+        if as_response and schema is None:
+            warnings.warn(
+                "as_response is deprecated and will be removed in v2.0. "
+                "Use schema=YourSchema instead.",
+                DeprecationWarning,
+                stacklevel=2,
+            )
         q = select(cls.model)
         if joins:
             for model, condition in joins:
@@ -214,8 +298,8 @@ class AsyncCrud(Generic[ModelType]):
                     else q.join(model, condition)
                 )
         q = q.where(and_(*filters))
-        if load_options:
+        if resolved := cls._resolve_load_options(load_options):
-            q = q.options(*load_options)
+            q = q.options(*resolved)
         if with_for_update:
             q = q.with_for_update()
         result = await session.execute(q)
@@ -223,8 +307,9 @@ class AsyncCrud(Generic[ModelType]):
         if not item:
             raise NotFoundError()
         result = cast(ModelType, item)
-        if as_response:
+        if as_response or schema:
-            return Response(data=result)
+            data_out = schema.model_validate(result) if schema else result
+            return Response(data=data_out)
         return result
 
     @classmethod
@@ -235,7 +320,7 @@ class AsyncCrud(Generic[ModelType]):
         *,
         joins: JoinType | None = None,
         outer_join: bool = False,
-        load_options: list[Any] | None = None,
+        load_options: list[ExecutableOption] | None = None,
     ) -> ModelType | None:
         """Get the first matching record, or None.
 
@@ -259,8 +344,8 @@ class AsyncCrud(Generic[ModelType]):
                 )
         if filters:
             q = q.where(and_(*filters))
-        if load_options:
+        if resolved := cls._resolve_load_options(load_options):
-            q = q.options(*load_options)
+            q = q.options(*resolved)
         result = await session.execute(q)
         return cast(ModelType | None, result.unique().scalars().first())
 
@@ -272,7 +357,7 @@ class AsyncCrud(Generic[ModelType]):
         filters: list[Any] | None = None,
         joins: JoinType | None = None,
         outer_join: bool = False,
-        load_options: list[Any] | None = None,
+        load_options: list[ExecutableOption] | None = None,
         order_by: Any | None = None,
         limit: int | None = None,
         offset: int | None = None,
@@ -302,8 +387,8 @@ class AsyncCrud(Generic[ModelType]):
                 )
         if filters:
             q = q.where(and_(*filters))
-        if load_options:
+        if resolved := cls._resolve_load_options(load_options):
-            q = q.options(*load_options)
+            q = q.options(*resolved)
         if order_by is not None:
             q = q.order_by(order_by)
         if offset is not None:
@@ -313,6 +398,21 @@ class AsyncCrud(Generic[ModelType]):
         result = await session.execute(q)
         return cast(Sequence[ModelType], result.unique().scalars().all())
 
+    @overload
+    @classmethod
+    async def update(  # pragma: no cover
+        cls: type[Self],
+        session: AsyncSession,
+        obj: BaseModel,
+        filters: list[Any],
+        *,
+        exclude_unset: bool = True,
+        exclude_none: bool = False,
+        schema: type[SchemaType],
+        as_response: bool = ...,
+    ) -> Response[SchemaType]: ...
+
+    # Backward-compatible - will be removed in v2.0
     @overload
     @classmethod
     async def update(  # pragma: no cover
@@ -324,6 +424,7 @@ class AsyncCrud(Generic[ModelType]):
         exclude_unset: bool = True,
         exclude_none: bool = False,
         as_response: Literal[True],
+        schema: None = ...,
     ) -> Response[ModelType]: ...
 
     @overload
@@ -337,6 +438,7 @@ class AsyncCrud(Generic[ModelType]):
         exclude_unset: bool = True,
         exclude_none: bool = False,
         as_response: Literal[False] = ...,
+        schema: None = ...,
     ) -> ModelType: ...
 
     @classmethod
@@ -349,7 +451,8 @@ class AsyncCrud(Generic[ModelType]):
         exclude_unset: bool = True,
         exclude_none: bool = False,
         as_response: bool = False,
-    ) -> ModelType | Response[ModelType]:
+        schema: type[BaseModel] | None = None,
+    ) -> ModelType | Response[ModelType] | Response[Any]:
         """Update a record in the database.
 
         Args:
@@ -358,20 +461,30 @@ class AsyncCrud(Generic[ModelType]):
             filters: List of SQLAlchemy filter conditions
             exclude_unset: Exclude fields not explicitly set in the schema
             exclude_none: Exclude fields with None value
-            as_response: If True, wrap result in Response object
+            as_response: Deprecated. Use ``schema`` instead. Will be removed in v2.0.
+            schema: Pydantic schema to serialize the result into. When provided,
+                the result is automatically wrapped in a ``Response[schema]``.
 
         Returns:
-            Updated model instance or Response wrapping it
+            Updated model instance, or ``Response[schema]`` when ``schema`` is given,
+            or ``Response[ModelType]`` when ``as_response=True`` (deprecated).
 
         Raises:
             NotFoundError: If no record found
         """
+        if as_response and schema is None:
+            warnings.warn(
+                "as_response is deprecated and will be removed in v2.0. "
+                "Use schema=YourSchema instead.",
+                DeprecationWarning,
+                stacklevel=2,
+            )
         async with get_transaction(session):
             m2m_exclude = cls._m2m_schema_fields()
 
             # Eagerly load M2M relationships that will be updated so that
             # setattr does not trigger a lazy load (which fails in async).
-            m2m_load_options: list[Any] = []
+            m2m_load_options: list[ExecutableOption] = []
             if m2m_exclude and cls.m2m_fields:
                 for schema_field, rel in cls.m2m_fields.items():
                     if schema_field in obj.model_fields_set:
@@ -395,8 +508,9 @@ class AsyncCrud(Generic[ModelType]):
                 for rel_attr, related_instances in m2m_resolved.items():
                     setattr(db_model, rel_attr, related_instances)
         await session.refresh(db_model)
-        if as_response:
+        if as_response or schema:
-            return Response(data=db_model)
+            data_out = schema.model_validate(db_model) if schema else db_model
+            return Response(data=data_out)
         return db_model
 
     @classmethod
@@ -478,11 +592,20 @@ class AsyncCrud(Generic[ModelType]):
         Args:
             session: DB async session
             filters: List of SQLAlchemy filter conditions
-            as_response: If True, wrap result in Response object
+            as_response: Deprecated. Will be removed in v2.0. When ``True``,
+                returns ``Response[None]`` instead of ``bool``.
 
         Returns:
-            True if deletion was executed, or Response wrapping it
+            ``True`` if deletion was executed, or ``Response[None]`` when
+            ``as_response=True`` (deprecated).
         """
+        if as_response:
+            warnings.warn(
+                "as_response is deprecated and will be removed in v2.0. "
+                "Use schema=YourSchema instead.",
+                DeprecationWarning,
+                stacklevel=2,
+            )
         async with get_transaction(session):
             q = sql_delete(cls.model).where(and_(*filters))
             await session.execute(q)
@@ -555,22 +678,60 @@ class AsyncCrud(Generic[ModelType]):
         result = await session.execute(q)
         return bool(result.scalar())
 
+    @overload
     @classmethod
-    async def paginate(
+    async def offset_paginate(  # pragma: no cover
         cls: type[Self],
         session: AsyncSession,
         *,
         filters: list[Any] | None = None,
         joins: JoinType | None = None,
         outer_join: bool = False,
-        load_options: list[Any] | None = None,
+        load_options: list[ExecutableOption] | None = None,
         order_by: Any | None = None,
         page: int = 1,
         items_per_page: int = 20,
         search: str | SearchConfig | None = None,
         search_fields: Sequence[SearchFieldType] | None = None,
-    ) -> PaginatedResponse[ModelType]:
+        schema: type[SchemaType],
-        """Get paginated results with metadata.
+    ) -> PaginatedResponse[SchemaType]: ...
+
+    # Backward-compatible - will be removed in v2.0
+    @overload
+    @classmethod
+    async def offset_paginate(  # pragma: no cover
+        cls: type[Self],
+        session: AsyncSession,
+        *,
+        filters: list[Any] | None = None,
+        joins: JoinType | None = None,
+        outer_join: bool = False,
+        load_options: list[ExecutableOption] | None = None,
+        order_by: Any | None = None,
+        page: int = 1,
+        items_per_page: int = 20,
+        search: str | SearchConfig | None = None,
+        search_fields: Sequence[SearchFieldType] | None = None,
+        schema: None = ...,
+    ) -> PaginatedResponse[ModelType]: ...
+
+    @classmethod
+    async def offset_paginate(
+        cls: type[Self],
+        session: AsyncSession,
+        *,
+        filters: list[Any] | None = None,
+        joins: JoinType | None = None,
+        outer_join: bool = False,
+        load_options: list[ExecutableOption] | None = None,
+        order_by: Any | None = None,
+        page: int = 1,
+        items_per_page: int = 20,
+        search: str | SearchConfig | None = None,
+        search_fields: Sequence[SearchFieldType] | None = None,
+        schema: type[BaseModel] | None = None,
+    ) -> PaginatedResponse[ModelType] | PaginatedResponse[Any]:
+        """Get paginated results using offset-based pagination.
 
         Args:
             session: DB async session
@@ -583,9 +744,10 @@ class AsyncCrud(Generic[ModelType]):
             items_per_page: Number of items per page
             search: Search query string or SearchConfig object
             search_fields: Fields to search in (overrides class default)
+            schema: Optional Pydantic schema to serialize each item into.
 
         Returns:
-            Dict with 'data' and 'pagination' keys
+            PaginatedResponse with OffsetPagination metadata
         """
         filters = list(filters) if filters else []
         offset = (page - 1) * items_per_page
@@ -619,14 +781,17 @@ class AsyncCrud(Generic[ModelType]):
 
         if filters:
             q = q.where(and_(*filters))
-        if load_options:
+        if resolved := cls._resolve_load_options(load_options):
-            q = q.options(*load_options)
+            q = q.options(*resolved)
         if order_by is not None:
             q = q.order_by(order_by)
 
         q = q.offset(offset).limit(items_per_page)
         result = await session.execute(q)
-        items = cast(list[ModelType], result.unique().scalars().all())
+        raw_items = cast(list[ModelType], result.unique().scalars().all())
+        items: list[Any] = (
+            [schema.model_validate(item) for item in raw_items] if schema else raw_items
+        )
 
         # Count query (with same joins and filters)
         pk_col = cls.model.__mapper__.primary_key[0]
@@ -654,7 +819,7 @@ class AsyncCrud(Generic[ModelType]):
 
         return PaginatedResponse(
             data=items,
-            pagination=Pagination(
+            pagination=OffsetPagination(
                 total_count=total_count,
                 items_per_page=items_per_page,
                 page=page,
@@ -662,12 +827,193 @@ class AsyncCrud(Generic[ModelType]):
             ),
         )
 
+    # Backward-compatible - will be removed in v2.0
+    paginate = offset_paginate
+
+    @overload
+    @classmethod
+    async def cursor_paginate(  # pragma: no cover
+        cls: type[Self],
+        session: AsyncSession,
+        *,
+        cursor: str | None = None,
+        filters: list[Any] | None = None,
+        joins: JoinType | None = None,
+        outer_join: bool = False,
+        load_options: list[ExecutableOption] | None = None,
+        order_by: Any | None = None,
+        items_per_page: int = 20,
+        search: str | SearchConfig | None = None,
+        search_fields: Sequence[SearchFieldType] | None = None,
+        schema: type[SchemaType],
+    ) -> PaginatedResponse[SchemaType]: ...
+
+    # Backward-compatible - will be removed in v2.0
+    @overload
+    @classmethod
+    async def cursor_paginate(  # pragma: no cover
+        cls: type[Self],
+        session: AsyncSession,
+        *,
+        cursor: str | None = None,
+        filters: list[Any] | None = None,
+        joins: JoinType | None = None,
+        outer_join: bool = False,
+        load_options: list[ExecutableOption] | None = None,
+        order_by: Any | None = None,
+        items_per_page: int = 20,
+        search: str | SearchConfig | None = None,
+        search_fields: Sequence[SearchFieldType] | None = None,
+        schema: None = ...,
+    ) -> PaginatedResponse[ModelType]: ...
+
+    @classmethod
+    async def cursor_paginate(
+        cls: type[Self],
+        session: AsyncSession,
+        *,
+        cursor: str | None = None,
+        filters: list[Any] | None = None,
+        joins: JoinType | None = None,
+        outer_join: bool = False,
+        load_options: list[ExecutableOption] | None = None,
+        order_by: Any | None = None,
+        items_per_page: int = 20,
+        search: str | SearchConfig | None = None,
+        search_fields: Sequence[SearchFieldType] | None = None,
+        schema: type[BaseModel] | None = None,
+    ) -> PaginatedResponse[ModelType] | PaginatedResponse[Any]:
+        """Get paginated results using cursor-based pagination.
+
+        Args:
+            session: DB async session.
+            cursor: Cursor string from a previous ``CursorPagination``.
+                Omit (or pass ``None``) to start from the beginning.
+            filters: List of SQLAlchemy filter conditions.
+            joins: List of ``(model, condition)`` tuples for joining related
+                tables.
+            outer_join: Use LEFT OUTER JOIN instead of INNER JOIN.
+            load_options: SQLAlchemy loader options. Falls back to
+                ``default_load_options`` when not provided.
+            order_by: Additional ordering applied after the cursor column.
+            items_per_page: Number of items per page (default 20).
+            search: Search query string or SearchConfig object.
+            search_fields: Fields to search in (overrides class default).
+            schema: Optional Pydantic schema to serialize each item into.
+
+        Returns:
+            PaginatedResponse with CursorPagination metadata
+        """
+        filters = list(filters) if filters else []
+        search_joins: list[Any] = []
+
+        if cls.cursor_column is None:
+            raise ValueError(
+                f"{cls.__name__}.cursor_column is not set. "
+                "Pass cursor_column=<column> to CrudFactory() to use cursor_paginate."
+            )
+        cursor_column: Any = cls.cursor_column
+        cursor_col_name: str = cursor_column.key
+
+        if cursor is not None:
+            raw_val = _decode_cursor(cursor)
+            col_type = cursor_column.property.columns[0].type
+            if isinstance(col_type, Integer):
+                cursor_val: Any = int(raw_val)
+            elif isinstance(col_type, Uuid):
+                cursor_val = uuid_module.UUID(raw_val)
+            elif isinstance(col_type, DateTime):
+                cursor_val = datetime.fromisoformat(raw_val)
+            elif isinstance(col_type, Date):
+                cursor_val = date.fromisoformat(raw_val)
+            elif isinstance(col_type, (Float, Numeric)):
+                cursor_val = Decimal(raw_val)
+            else:
+                raise ValueError(
+                    f"Unsupported cursor column type: {type(col_type).__name__!r}. "
+                    "Supported types: Integer, BigInteger, SmallInteger, Uuid, "
+                    "DateTime, Date, Float, Numeric."
+                )
+            filters.append(cursor_column > cursor_val)
+
+        # Build search filters
+        if search:
+            search_filters, search_joins = build_search_filters(
+                cls.model,
+                search,
+                search_fields=search_fields,
+                default_fields=cls.searchable_fields,
+            )
+            filters.extend(search_filters)
+
+        # Build query
+        q = select(cls.model)
+
+        # Apply explicit joins
+        if joins:
+            for model, condition in joins:
+                q = (
+                    q.outerjoin(model, condition)
+                    if outer_join
+                    else q.join(model, condition)
+                )
+
+        # Apply search joins (always outer joins)
+        for join_rel in search_joins:
+            q = q.outerjoin(join_rel)
+
+        if filters:
+            q = q.where(and_(*filters))
+        if resolved := cls._resolve_load_options(load_options):
+            q = q.options(*resolved)
+
+        # Cursor column is always the primary sort
+        q = q.order_by(cursor_column)
+        if order_by is not None:
+            q = q.order_by(order_by)
+
+        # Fetch one extra to detect whether a next page exists
+        q = q.limit(items_per_page + 1)
+        result = await session.execute(q)
+        raw_items = cast(list[ModelType], result.unique().scalars().all())
+
+        has_more = len(raw_items) > items_per_page
+        items_page = raw_items[:items_per_page]
+
+        # next_cursor points past the last item on this page
+        next_cursor: str | None = None
+        if has_more and items_page:
+            next_cursor = _encode_cursor(getattr(items_page[-1], cursor_col_name))
+
+        # prev_cursor points to the first item on this page or None when on the first page
+        prev_cursor: str | None = None
+        if cursor is not None and items_page:
+            prev_cursor = _encode_cursor(getattr(items_page[0], cursor_col_name))
+
+        items: list[Any] = (
+            [schema.model_validate(item) for item in items_page]
+            if schema
+            else items_page
+        )
+
+        return PaginatedResponse(
+            data=items,
+            pagination=CursorPagination(
+                next_cursor=next_cursor,
+                prev_cursor=prev_cursor,
+                items_per_page=items_per_page,
+                has_more=has_more,
+            ),
+        )
+
 
 def CrudFactory(
     model: type[ModelType],
     *,
     searchable_fields: Sequence[SearchFieldType] | None = None,
     m2m_fields: M2MFieldType | None = None,
+    default_load_options: list[ExecutableOption] | None = None,
+    cursor_column: Any | None = None,
 ) -> type[AsyncCrud[ModelType]]:
     """Create a CRUD class for a specific model.
 
@@ -677,6 +1023,14 @@ def CrudFactory(
         m2m_fields: Optional mapping for many-to-many relationships.
             Maps schema field names (containing lists of IDs) to
             SQLAlchemy relationship attributes.
+        default_load_options: Default SQLAlchemy loader options applied to all read
+            queries when no explicit ``load_options`` are passed. Use this
+            instead of ``lazy="selectin"`` on the model so that loading
+            strategy is explicit and per-CRUD. Overridden entirely (not
+            merged) when ``load_options`` is provided at call-site.
+        cursor_column: Required to call ``cursor_paginate``.
+            Must be monotonically ordered (e.g. integer PK, UUID v7, timestamp).
+            See the cursor pagination docs for supported column types.
 
     Returns:
         AsyncCrud subclass bound to the model
@@ -702,6 +1056,25 @@ def CrudFactory(
             m2m_fields={"tag_ids": Post.tags},
         )
 
+        # With a fixed cursor column for cursor_paginate:
+        PostCrud = CrudFactory(
+            Post,
+            cursor_column=Post.created_at,
+        )
+
+        # With default load strategy (replaces lazy="selectin" on the model):
+        ArticleCrud = CrudFactory(
+            Article,
+            default_load_options=[selectinload(Article.category), selectinload(Article.tags)],
+        )
+
+        # Override default_load_options for a specific call:
+        article = await ArticleCrud.get(
+            session,
+            [Article.id == 1],
+            load_options=[selectinload(Article.category)],  # tags won't load
+        )
+
         # Usage
         user = await UserCrud.get(session, [User.id == 1])
         posts = await PostCrud.get_multi(session, filters=[Post.user_id == user.id])
@@ -710,7 +1083,7 @@ def CrudFactory(
         post = await PostCrud.create(session, PostCreate(title="Hello", tag_ids=[id1, id2]))
 
         # With search
-        result = await UserCrud.paginate(session, search="john")
+        result = await UserCrud.offset_paginate(session, search="john")
 
         # With joins (inner join by default):
         users = await UserCrud.get_multi(
@@ -734,6 +1107,8 @@ def CrudFactory(
             "model": model,
             "searchable_fields": searchable_fields,
             "m2m_fields": m2m_fields,
+            "default_load_options": default_load_options,
+            "cursor_column": cursor_column,
         },
     )
     return cast(type[AsyncCrud[ModelType]], cls)

src/fastapi_toolsets/schemas.py

@@ -7,7 +7,9 @@ from pydantic import BaseModel, ConfigDict
 
 __all__ = [
     "ApiError",
+    "CursorPagination",
     "ErrorResponse",
+    "OffsetPagination",
     "Pagination",
     "PaginatedResponse",
     "PydanticBase",
@@ -90,8 +92,8 @@ class ErrorResponse(BaseResponse):
     data: Any | None = None
 
 
-class Pagination(PydanticBase):
+class OffsetPagination(PydanticBase):
-    """Pagination metadata for list responses.
+    """Pagination metadata for offset-based list responses.
 
     Attributes:
         total_count: Total number of items across all pages
@@ -106,17 +108,28 @@ class Pagination(PydanticBase):
     has_more: bool
 
 
-class PaginatedResponse(BaseResponse, Generic[DataT]):
+# Backward-compatible - will be removed in v2.0
-    """Paginated API response for list endpoints.
+Pagination = OffsetPagination
 
-    Example:
+
-        ```python
+class CursorPagination(PydanticBase):
-        PaginatedResponse[UserRead](
+    """Pagination metadata for cursor-based list responses.
-            data=users,
+
-            pagination=Pagination(total_count=100, items_per_page=10, page=1, has_more=True)
+    Attributes:
-        )
+        next_cursor: Encoded cursor for the next page, or None on the last page.
-        ```
+        prev_cursor: Encoded cursor for the previous page, or None on the first page.
+        items_per_page: Number of items requested per page.
+        has_more: Whether there is at least one more page after this one.
     """
 
+    next_cursor: str | None
+    prev_cursor: str | None = None
+    items_per_page: int
+    has_more: bool
+
+
+class PaginatedResponse(BaseResponse, Generic[DataT]):
+    """Paginated API response for list endpoints."""
+
     data: list[DataT]
-    pagination: Pagination
+    pagination: OffsetPagination | CursorPagination

tests/conftest.py

@@ -5,11 +5,25 @@ import uuid
 
 import pytest
 from pydantic import BaseModel
-from sqlalchemy import Column, ForeignKey, String, Table, Uuid
+import datetime
+import decimal
+
+from sqlalchemy import (
+    Column,
+    Date,
+    DateTime,
+    ForeignKey,
+    Integer,
+    Numeric,
+    String,
+    Table,
+    Uuid,
+)
 from sqlalchemy.ext.asyncio import async_sessionmaker, create_async_engine
 from sqlalchemy.orm import DeclarativeBase, Mapped, mapped_column, relationship
 
 from fastapi_toolsets.crud import CrudFactory
+from fastapi_toolsets.schemas import PydanticBase
 
 DATABASE_URL = os.getenv(
     key="DATABASE_URL",
@@ -69,6 +83,36 @@ post_tags = Table(
 )
 
 
+class IntRole(Base):
+    """Test role model with auto-increment integer PK."""
+
+    __tablename__ = "int_roles"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    name: Mapped[str] = mapped_column(String(50), unique=True)
+
+
+class Event(Base):
+    """Test model with DateTime and Date cursor columns."""
+
+    __tablename__ = "events"
+
+    id: Mapped[uuid.UUID] = mapped_column(Uuid, primary_key=True, default=uuid.uuid4)
+    name: Mapped[str] = mapped_column(String(100))
+    occurred_at: Mapped[datetime.datetime] = mapped_column(DateTime)
+    scheduled_date: Mapped[datetime.date] = mapped_column(Date)
+
+
+class Product(Base):
+    """Test model with Numeric cursor column."""
+
+    __tablename__ = "products"
+
+    id: Mapped[uuid.UUID] = mapped_column(Uuid, primary_key=True, default=uuid.uuid4)
+    name: Mapped[str] = mapped_column(String(100))
+    price: Mapped[decimal.Decimal] = mapped_column(Numeric(10, 2))
+
+
 class Post(Base):
     """Test post model."""
 
@@ -90,6 +134,13 @@ class RoleCreate(BaseModel):
     name: str
 
 
+class RoleRead(PydanticBase):
+    """Schema for reading a role."""
+
+    id: uuid.UUID
+    name: str
+
+
 class RoleUpdate(BaseModel):
     """Schema for updating a role."""
 
@@ -106,6 +157,13 @@ class UserCreate(BaseModel):
     role_id: uuid.UUID | None = None
 
 
+class UserRead(PydanticBase):
+    """Schema for reading a user (subset of fields — no email)."""
+
+    id: uuid.UUID
+    username: str
+
+
 class UserUpdate(BaseModel):
     """Schema for updating a user."""
 
@@ -160,11 +218,40 @@ class PostM2MUpdate(BaseModel):
     tag_ids: list[uuid.UUID] | None = None
 
 
+class IntRoleCreate(BaseModel):
+    """Schema for creating an IntRole."""
+
+    name: str
+
+
+class EventCreate(BaseModel):
+    """Schema for creating an Event."""
+
+    name: str
+    occurred_at: datetime.datetime
+    scheduled_date: datetime.date
+
+
+class ProductCreate(BaseModel):
+    """Schema for creating a Product."""
+
+    name: str
+    price: decimal.Decimal
+
+
 RoleCrud = CrudFactory(Role)
+RoleCursorCrud = CrudFactory(Role, cursor_column=Role.id)
+IntRoleCursorCrud = CrudFactory(IntRole, cursor_column=IntRole.id)
 UserCrud = CrudFactory(User)
+UserCursorCrud = CrudFactory(User, cursor_column=User.id)
 PostCrud = CrudFactory(Post)
 TagCrud = CrudFactory(Tag)
 PostM2MCrud = CrudFactory(Post, m2m_fields={"tag_ids": Post.tags})
+EventCrud = CrudFactory(Event)
+EventDateTimeCursorCrud = CrudFactory(Event, cursor_column=Event.occurred_at)
+EventDateCursorCrud = CrudFactory(Event, cursor_column=Event.scheduled_date)
+ProductCrud = CrudFactory(Product)
+ProductNumericCursorCrud = CrudFactory(Product, cursor_column=Product.price)
 
 
 @pytest.fixture

tests/test_crud_search.py

@@ -6,6 +6,7 @@ import pytest
 from sqlalchemy.ext.asyncio import AsyncSession
 
 from fastapi_toolsets.crud import SearchConfig, get_searchable_fields
+from fastapi_toolsets.schemas import OffsetPagination
 
 from .conftest import (
     Role,
@@ -39,6 +40,7 @@ class TestPaginateSearch:
             search_fields=[User.username],
         )
 
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 2
 
     @pytest.mark.anyio
@@ -57,6 +59,7 @@ class TestPaginateSearch:
             search_fields=[User.username, User.email],
         )
 
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 2
 
     @pytest.mark.anyio
@@ -84,6 +87,7 @@ class TestPaginateSearch:
             search_fields=[(User.role, Role.name)],
         )
 
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 2
 
     @pytest.mark.anyio
@@ -102,6 +106,7 @@ class TestPaginateSearch:
             search_fields=[User.username, (User.role, Role.name)],
         )
 
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 1
 
     @pytest.mark.anyio
@@ -117,6 +122,7 @@ class TestPaginateSearch:
             search_fields=[User.username],
         )
 
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 1
 
     @pytest.mark.anyio
@@ -132,6 +138,7 @@ class TestPaginateSearch:
             search=SearchConfig(query="johndoe", case_sensitive=True),
             search_fields=[User.username],
         )
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 0
 
         # Should find (case match)
@@ -140,6 +147,7 @@ class TestPaginateSearch:
             search=SearchConfig(query="JohnDoe", case_sensitive=True),
             search_fields=[User.username],
         )
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 1
 
     @pytest.mark.anyio
@@ -153,9 +161,11 @@ class TestPaginateSearch:
         )
 
         result = await UserCrud.paginate(db_session, search="")
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 2
 
         result = await UserCrud.paginate(db_session, search=None)
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 2
 
     @pytest.mark.anyio
@@ -177,6 +187,7 @@ class TestPaginateSearch:
             search_fields=[User.username],
         )
 
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 1
         assert result.data[0].username == "active_john"
 
@@ -189,6 +200,7 @@ class TestPaginateSearch:
 
         result = await UserCrud.paginate(db_session, search="findme")
 
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 1
 
     @pytest.mark.anyio
@@ -204,6 +216,7 @@ class TestPaginateSearch:
             search_fields=[User.username],
         )
 
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 0
         assert result.data == []
 
@@ -224,6 +237,7 @@ class TestPaginateSearch:
             items_per_page=5,
         )
 
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 15
         assert len(result.data) == 5
         assert result.pagination.has_more is True
@@ -248,6 +262,7 @@ class TestPaginateSearch:
             search_fields=[User.username],
         )
 
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 2
 
     @pytest.mark.anyio
@@ -270,6 +285,7 @@ class TestPaginateSearch:
             order_by=User.username,
         )
 
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 3
         usernames = [u.username for u in result.data]
         assert usernames == ["alice", "bob", "charlie"]
@@ -292,6 +308,7 @@ class TestPaginateSearch:
             search_fields=[User.id, User.username],
         )
 
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 1
         assert result.data[0].id == user_id
 
@@ -318,6 +335,7 @@ class TestSearchConfig:
             search_fields=[User.username, User.email],
         )
 
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 1
         assert result.data[0].username == "john_test"
 
@@ -333,6 +351,7 @@ class TestSearchConfig:
             search=SearchConfig(query="findme", fields=[User.email]),
         )
 
+        assert isinstance(result.pagination, OffsetPagination)
         assert result.pagination.total_count == 1
 
 

tests/test_schemas.py

@@ -5,7 +5,9 @@ from pydantic import ValidationError
 
 from fastapi_toolsets.schemas import (
     ApiError,
+    CursorPagination,
     ErrorResponse,
+    OffsetPagination,
     PaginatedResponse,
     Pagination,
     Response,
@@ -154,12 +156,12 @@ class TestErrorResponse:
         assert data["description"] == "Details"
 
 
-class TestPagination:
+class TestOffsetPagination:
-    """Tests for Pagination schema."""
+    """Tests for OffsetPagination schema (canonical name for offset-based pagination)."""
 
     def test_create_pagination(self):
-        """Create Pagination with all fields."""
+        """Create OffsetPagination with all fields."""
-        pagination = Pagination(
+        pagination = OffsetPagination(
             total_count=100,
             items_per_page=10,
             page=1,
@@ -173,7 +175,7 @@ class TestPagination:
 
     def test_last_page_has_more_false(self):
         """Last page has has_more=False."""
-        pagination = Pagination(
+        pagination = OffsetPagination(
             total_count=25,
             items_per_page=10,
             page=3,
@@ -183,8 +185,8 @@ class TestPagination:
         assert pagination.has_more is False
 
     def test_serialization(self):
-        """Pagination serializes correctly."""
+        """OffsetPagination serializes correctly."""
-        pagination = Pagination(
+        pagination = OffsetPagination(
             total_count=50,
             items_per_page=20,
             page=2,
@@ -197,6 +199,77 @@ class TestPagination:
         assert data["page"] == 2
         assert data["has_more"] is True
 
+    def test_pagination_alias_is_offset_pagination(self):
+        """Pagination is a backward-compatible alias for OffsetPagination."""
+        assert Pagination is OffsetPagination
+
+    def test_pagination_alias_constructs_offset_pagination(self):
+        """Code using Pagination(...) still works unchanged."""
+        pagination = Pagination(
+            total_count=10,
+            items_per_page=5,
+            page=2,
+            has_more=False,
+        )
+        assert isinstance(pagination, OffsetPagination)
+
+
+class TestCursorPagination:
+    """Tests for CursorPagination schema."""
+
+    def test_create_with_next_cursor(self):
+        """CursorPagination with a next cursor indicates more pages."""
+        pagination = CursorPagination(
+            next_cursor="eyJ2YWx1ZSI6ICIxMjMifQ==",
+            items_per_page=20,
+            has_more=True,
+        )
+        assert pagination.next_cursor == "eyJ2YWx1ZSI6ICIxMjMifQ=="
+        assert pagination.prev_cursor is None
+        assert pagination.items_per_page == 20
+        assert pagination.has_more is True
+
+    def test_create_last_page(self):
+        """CursorPagination for the last page has next_cursor=None and has_more=False."""
+        pagination = CursorPagination(
+            next_cursor=None,
+            items_per_page=20,
+            has_more=False,
+        )
+        assert pagination.next_cursor is None
+        assert pagination.has_more is False
+
+    def test_prev_cursor_defaults_to_none(self):
+        """prev_cursor defaults to None."""
+        pagination = CursorPagination(
+            next_cursor=None, items_per_page=10, has_more=False
+        )
+        assert pagination.prev_cursor is None
+
+    def test_prev_cursor_can_be_set(self):
+        """prev_cursor can be explicitly set."""
+        pagination = CursorPagination(
+            next_cursor="next123",
+            prev_cursor="prev456",
+            items_per_page=10,
+            has_more=True,
+        )
+        assert pagination.prev_cursor == "prev456"
+
+    def test_serialization(self):
+        """CursorPagination serializes correctly."""
+        pagination = CursorPagination(
+            next_cursor="abc123",
+            prev_cursor="xyz789",
+            items_per_page=20,
+            has_more=True,
+        )
+        data = pagination.model_dump()
+        assert data["next_cursor"] == "abc123"
+        assert data["prev_cursor"] == "xyz789"
+        assert data["items_per_page"] == 20
+        assert data["has_more"] is True
+
 
 class TestPaginatedResponse:
     """Tests for PaginatedResponse schema."""
@@ -214,6 +287,7 @@ class TestPaginatedResponse:
             pagination=pagination,
         )
 
+        assert isinstance(response.pagination, OffsetPagination)
         assert len(response.data) == 2
         assert response.pagination.total_count == 30
         assert response.status == ResponseStatus.SUCCESS
@@ -247,6 +321,7 @@ class TestPaginatedResponse:
             pagination=pagination,
         )
 
+        assert isinstance(response.pagination, OffsetPagination)
         assert response.data == []
         assert response.pagination.total_count == 0
 
@@ -290,6 +365,36 @@ class TestPaginatedResponse:
         assert data["data"] == ["item1", "item2"]
         assert data["pagination"]["page"] == 5
 
+    def test_pagination_field_accepts_offset_pagination(self):
+        """PaginatedResponse.pagination accepts OffsetPagination."""
+        response = PaginatedResponse(
+            data=[1, 2],
+            pagination=OffsetPagination(
+                total_count=2, items_per_page=10, page=1, has_more=False
+            ),
+        )
+        assert isinstance(response.pagination, OffsetPagination)
+
+    def test_pagination_field_accepts_cursor_pagination(self):
+        """PaginatedResponse.pagination accepts CursorPagination."""
+        response = PaginatedResponse(
+            data=[1, 2],
+            pagination=CursorPagination(
+                next_cursor=None, items_per_page=10, has_more=False
+            ),
+        )
+        assert isinstance(response.pagination, CursorPagination)
+
+    def test_pagination_alias_accepted(self):
+        """Constructing PaginatedResponse with Pagination (alias) still works."""
+        response = PaginatedResponse(
+            data=[],
+            pagination=Pagination(
+                total_count=0, items_per_page=10, page=1, has_more=False
+            ),
+        )
+        assert isinstance(response.pagination, OffsetPagination)
+
 
 class TestFromAttributes:
     """Tests for from_attributes config (ORM mode)."""

uv.lock

@@ -251,7 +251,7 @@ wheels = [
 
 [[package]]
 name = "fastapi-toolsets"
-version = "0.10.0"
+version = "1.1.2"
 source = { editable = "." }
 dependencies = [
     { name = "asyncpg" },