Skip to content

Overview

AutoCRUD is a model-driven backend framework for FastAPI that manages application data as versioned resources.

Instead of manually implementing CRUD APIs, validation, search, and storage logic, AutoCRUD generates these capabilities automatically from your resource models.

The core idea

AutoCRUD manages data using three fundamental concepts:

  • Resource
  • Revision
  • ResourceManager

Together they form the foundation of the framework.

Resource

A resource represents a logical entity in your application.

Examples:

  • User
  • Article
  • Job
  • Configuration

Resources are defined as Python data models.

class User(Struct):
    name: str
    email: str

Each resource has a unique identifier (resource_id) and associated metadata.

Revision

AutoCRUD stores data as immutable revisions.

Instead of overwriting data, updates create new revisions.

Resource
 ├── r1
 ├── r2
 └── r3

The currently active revision is tracked by metadata:

ResourceMeta.current_revision_id

This enables built-in support for:

  • audit history
  • rollback
  • draft workflows
  • debugging

ResourceManager

All operations on resources go through the ResourceManager.

Example:

resource_manager.create(data)
resource_manager.get(resource_id)
resource_manager.update(resource_id, new_data)

The manager coordinates:

  • validation
  • revision creation
  • metadata updates
  • storage operations
  • event handlers

Developers do not interact directly with storage systems.

Schema and validation

Resources may define schemas that control validation and constraints.

Examples include:

  • field validation
  • uniqueness constraints
  • custom validation hooks

These checks run automatically when resources are created or updated.

References

Resources and revisions can be referenced using:

  • Ref – reference to a resource
  • RefRevision – reference to a specific revision

These types allow relationships between resources while preserving revision history.

Summary

AutoCRUD is built around a simple mental model:

ResourceManager
    ↓
Resource
    ↓
Revisions

By combining versioned resources, metadata indexing, and storage abstraction, AutoCRUD provides a consistent way to build APIs without writing repetitive infrastructure code.