LinaPro is an AI-native full-stack framework built for sustainable delivery. It brings together a specification-driven AI development workflow, a comprehensive AI skill system spanning the entire development lifecycle, a complete plugin runtime, and an integrated full-stack design — with enterprise-grade capabilities like access control, system configuration, and job scheduling built right in.

Teams can skip the infrastructure-from-scratch phase and put AI to work driving real business development from day one.

LinaPro is designed for individual developers, engineering teams, and enterprises. Here's what it brings to the table:

• AI-native development workflow: Ships with a specification-driven AI development workflow, with first-class support for the optional but recommended OpenSpec tool. AI leads analysis, design, and implementation while every change is anchored to incremental specs and mandatory E2E tests — freeing your team to focus on direction rather than execution details.
• A rich AI skill ecosystem: Over a dozen built-in AI skills cover the full development lifecycle — backend development, frontend design, test writing, code review, performance auditing, version upgrades, and more. AI makes framework-aware decisions in every context without needing to be re-briefed each session.
• Fast business development: A batteries-included management workspace and a rich set of built-in modules dramatically shorten the path from zero to production.
• Integrated full-stack design: Frontend and backend are designed as a unified whole — API contracts, permission models, and design conventions are fully aligned, with no manual cross-framework integration overhead.
• Complete API documentation: All host and plugin API endpoints are automatically aggregated and exposed as a single browsable, debuggable doc site.
• Extensible plugin ecosystem: A dual-mode plugin system — source plugins and WASM dynamic plugins — lets any capability be extended or replaced. Official plugins are maintained as a separate submodule and pulled in only when needed, keeping the core framework lean.
• Multi-tenant support: Native multi-tenant capability with an official multi-tenant management plugin. When the plugin is not enabled, the system automatically falls back to single-tenant mode with zero migration cost.
• Enterprise-grade governance: JWT authentication paired with a declarative RBAC permission system, plus built-in operation logs, login logs, and session management for comprehensive auditability.
• Distribution-ready by design: Built-in distributed locking, key-value caching, and horizontal scaling. Cluster mode is coordinated via Redis for high availability — no changes to business code required.

graph TB
 subgraph Workflow["AI Development Workflow openspec/"]
 direction LR
 Explore["🔍 Explore"] --> Propose["📋 Propose"] --> Implement["⚙️ Implement"] --> Review["🔎 Review"] --> Archive["📦 Archive"]
 end
 subgraph Frontend["Management Workspace lina-vben"]
 UI["Vue 3 + Vben5 + Ant Design"]
 end
 subgraph Host["Core Host Service lina-core"]
 direction TB
 API["API Layer\n(g.Meta route definitions + DTO)"]
 Ctrl["Controller Layer\n(HTTP request handling)"]
 Svc["Service Layer\n(core business logic)"]
 Plugin["Plugin Runtime\n(lifecycle orchestration · sandbox isolation)"]
 Tenant["Native Multi-Tenant\n(bizctx · tenant_id)"]
 Gov["Governance\n(JWT · RBAC · Logs · Sessions)"]
 API --> Ctrl --> Svc
 Svc --> Plugin
 Svc --> Tenant
 Svc --> Gov
 end
 subgraph Plugins["Plugin System apps/lina-plugins"]
 direction LR
 Source["Source Plugins\ncompiled with host"]
 Dynamic["WASM Dynamic Plugins\nhot-loaded at runtime"]
 end
 DB[("Data Store\nPostgreSQL")]
 Redis[("Cluster Coordination\nRedis")]
 Workflow -.->|spec-driven| Frontend
 Workflow -.->|spec-driven| Host
 UI -->|HTTP| API
 Plugin -->|compiled load| Source
 Plugin -->|sandbox execution| Dynamic
 Svc --> DB
 Gov --> DB
 Svc -.->|cluster.enabled=true| Redis