ObjectStack
The AI-native application framework behind the stack
ObjectStack is the open-source framework for metadata-defined business applications: objects, permissions, flows, views, actions, APIs, SDKs, and AI tools. ObjectQL becomes an implementation detail here — the internal data/query engine and driver contract — while the public developer experience centers on ObjectStack docs and the framework repository.
From ObjectStack metadata to runtime surfaces
ObjectStack models the business; ObjectOS and ObjectUI turn it into an AI-operable application.
// customer.object.ts
import { defineObject, z } from '@objectstack/spec';
export const Customer = defineObject({
name: 'customer',
label: 'Customer',
fields: {
companyName: z.string().min(1),
accountManagerId: z.lookup('user'),
annualRevenue: z.number().optional(),
},
});// Derived by the ObjectStack runtime:
// Data engine -> driver-backed CRUD/query
// REST API -> /api/v1/data/customer
// Client SDK -> client.data.find('customer', ...)
// ObjectUI -> forms, tables, dashboards
// MCP tools -> customer.list / customer.update
// ObjectOS enforces auth, row/field permissions,
// validation, audit, and request attribution.What Lives in ObjectStack
AI-ready metadata first; data/query plumbing behind the scenes.
Business Metadata Core
ObjectStack defines objects, fields, permissions, flows, actions, views, agents, and tool surfaces in one structured model. The runtime decides who can use them and how each action is exposed.
Driver-Based Storage
Run the same object metadata through SQL, MongoDB, memory, spreadsheets, or custom drivers. The internal data/query engine serves the framework instead of becoming the product story.
Readable by Humans and AI
Objects, actions, and query surfaces are structured enough for the framework, REST layer, SDKs, ObjectUI, and MCP tools to derive useful behavior without duplicate definitions.
ObjectStack vs. ORM-Only Backends
ObjectStack is the metadata spine for apps, APIs, UI, and AI tools; the data/query engine stays internal.
Open-source framework for business metadata, APIs, UI, and AI tool surfaces.
Usually only a data abstraction, forcing custom glue around it.
ObjectOS applies auth, RBAC/RLS/FLS, audit, and agent identity around every action.
Governance must be rebuilt per API route and per AI integration.
Metadata can move across drivers and runtimes as a compiled artifact.
Models often bind to a dialect, migration stack, or application framework.
Actions and object metadata feed MCP tools through the runtime.
Agent tools are usually hand-written and drift from the backend.
Drivers Are the Data Boundary
ObjectStack owns the common object/action contract; drivers own the actual storage behavior.
PostgreSQL
Production relational storage through the SQL driver.
MySQL
Relational storage for common web and enterprise deployments.
SQLite
Local development, embedded tools, and standalone deployments.
SQL Server
Enterprise relational systems via custom or commercial drivers.
Oracle Database
Legacy enterprise systems can be mounted without immediate migration.
MongoDB
Document storage through native driver semantics where appropriate.
Excel / CSV
Spreadsheets can be wrapped as governed objects for APIs, UI, and agents.
Use Cases
Use ObjectStack as the open-source foundation; let ObjectOS run it and ObjectUI render it.
AI-Ready App Models
Prototype objects, flows, actions, and agent tools locally; move the same metadata into ObjectOS when you need auth, audit, APIs, and governed execution.
Driver Adapters
Wrap existing databases, spreadsheets, and proprietary systems behind ObjectStack metadata so APIs, UI, and AI tools can reason about them safely.
Product Metadata Spine
Provide consistent filters, pagination, actions, permissions, and tool contracts for generated REST APIs, client SDKs, UI renderers, and AI agents.
Define Your Business Once.
Let Humans And Agents Operate It Safely.
Stop scattering business logic across SQL, JavaScript, and UI state. Make objects, permissions, workflows, APIs, UI metadata, and agent tools one structured Zod source of truth — analyzable, versioned, and auditable end to end. Launch July 2026.