Basic product features
Lovrabet is an AI-Native platform for building and running intelligent enterprise business systems. Its capabilities fall into two main categories: the Lovrabet app admin console, for building, configuring, and governing apps and managing their assets; and the Lovrabet runtime app system, for the end users who run the business. On top of these, the Lovrabet CLI and RabetBase CLI open the platform up to enterprise systems, Agent tools, and intelligent development tools.
Capability overview
| Category | Audience | Core value |
|---|---|---|
| Lovrabet app admin console | Product, operations, implementation, developers, admins | Build apps, manage AI assets, configure permissions, generate pages, connect databases, and accumulate data and business capabilities |
| Lovrabet runtime app system | Business users, frontline staff, managers | Conversational Agent as the default entry point, plus classic standard system pages for queries, data entry, approvals, and analysis |
| CLI | RabetBase CLI: developers, Agent tools, enterprise integrators Lovrabet CLI: business users, frontline staff, managers | Let external intelligent tools, enterprise systems, and dev environments call Lovrabet app capabilities directly |
1. Lovrabet app admin console
The app admin console covers the full journey of an enterprise app — creation, configuration, page development, data modeling, AI asset accumulation, and permission governance. It serves both classic system building and the production and governance of AI Agent capabilities.
1. App management
App management handles app-level configuration, operational insights, AI asset accumulation, and security governance — it is the control center of an enterprise app.
App insights
App insights show how well an app is built, how deeply AI is used, and how the team collaborates. It's not simple traffic statistics — it helps a team judge whether the app has truly moved into an AI-Native way of working.
| Dimension | Description |
|---|---|
| AI-Native maturity | Gauges how intelligent the app is, via AI coverage, active adoption, deep workflow rate, top performers, and similar metrics |
| AI tool usage | Tracks usage of Agents, CLI, automation bindings, development actions, and more |
| App asset statistics | Monitors accumulated assets such as pages, menus, datasets, Backend Functions, and custom SQL |
| Team member performance | See how each member performs in AI usage, development actions, asset contributions, and deep workflows |
| Next-step suggestions | Improvement recommendations based on the app's current state — for example, replicating top performers' practices or filling in deep workflows |
App insights are designed for admins, product owners, and delivery leads, to tell whether an app merely "has generated pages" or has genuinely embedded AI into daily workflows.
AI asset management
AI asset management governs the enterprise intelligence assets an Agent can call, learn from, and execute. Enterprise know-how stops living only in people, documents, or project hand-offs and becomes capability the system can reuse.
| Asset type | Description |
|---|---|
| Rules | Rule assets that constrain Agent behavior, business judgment, execution boundaries, and security policies |
| Skills | Skill assets that package high-frequency business processes, automated actions, and role-specific SOPs into callable capabilities |
| Knowledges | Knowledge assets that capture business descriptions, product material, process standards, FAQs, and industry knowledge |
The core purpose of AI asset management is turning business rules, process experience, and knowledge material into assets an Agent can understand, call, and keep improving. Things like converting a lead to an opportunity, changing customer lifecycle states, or updating upstream records once an order is signed can all be captured as Skills.
App basics configuration
App basics configuration defines the app's identity in both the console and at runtime.
| Setting | Description |
|---|---|
| App name & description | Define the app's name, business summary, and use cases so users understand what it's for |
| App logo | Configure the app's icon, for the enterprise portal and runtime display |
| Multilingual settings | Show different names and content to users in different languages |
| Publish status | Toggle the app between unpublished and published, to separate the build phase from live operation |
| Domain configuration | Set the app's access domain or a custom domain to match enterprise entry-point requirements |
| Theme & advanced settings | Configure themes, UI style, and advanced parameters so the app matches the company's brand and habits |
These settings solve the app's "identity" problem — giving a system generated by AI or extended by developers the baseline configuration of a proper enterprise app.
App permission management
App permission management controls who can enter the app, which pages they see, which components they can operate, and which data they can access. It is a key security foundation for running Lovrabet apps in the enterprise.
| Managed object | Description |
|---|---|
| Roles | Create roles such as admin, developer, and regular user, and assign in-app permissions to each |
| Members | Manage app members: onboarding, role assignment, and access scope |
| Page permissions | Control which menus, pages, and modules each role can access |
| Component permissions | Control component-level capabilities such as buttons, fields, table actions, and bulk operations |
| Data scope | Restrict visible data by role, department, organization, owner, region, and other dimensions |
| High-risk action control | Apply permission and audit controls to deletes, exports, bulk updates, and sensitive data access |
With permission management, one app can serve managers, business users, operations staff, and external collaborators at the same time — each seeing exactly the content and actions their role permits.
Notification configuration
Notification configuration manages how the app delivers messages and business reminders.
| Capability | Description |
|---|---|
| Channel management | Configure the channels the app can use — enterprise IM, system messages, email, or other extensions |
| Business reminders | Send notifications for approvals, anomalies, tasks, status changes, and similar events |
| Agent collaboration alerts | After spotting a risk, producing a suggestion, or finishing a task, the Agent can notify the people involved through the configured channels |
| Process notifications | Reach the owner at key process milestones to keep tasks moving to completion |
Notifications turn the app from a passive page users must check into one that proactively pushes key changes, risks, and to-dos to the right roles.
App asset management
App asset management centrally manages everything produced while building the app: pages, data, APIs, SQL, BFFs, components, and AI assets.
| Asset type | Description |
|---|---|
| Page assets | Manage list pages, detail pages, create pages, edit pages, dashboards, and more |
| Data assets | Manage database connections, tables, datasets, custom SQL, and the business data domain model |
| Backend assets | Manage Backend Functions, Open APIs, business actions, and API extensions |
| Component assets | Manage page components, business components, third-party components, and reusable interaction units |
| AI assets | Manage Rules, Skills, Knowledges, and the tool capabilities an Agent can call |
The value of app asset management is that building an app is no longer one-off page generation — assets keep accumulating as reusable, governable, extensible enterprise property.
2. Page development
Page development generates, manages, and edits app pages. Lovrabet can generate system pages from database structures, business requirements, or Agent conversations, and every page stays open to component-level editing and permission control afterwards.
Menu management
Menu management organizes the menus, modules, and page entries of the runtime app.
| Capability | Description |
|---|---|
| Menu structure | Organize modules by business domain: customer management, sales management, product management, transaction management, organization management, and so on |
| Page grouping | Mount list, detail, create, edit, and dashboard pages onto the right menu entries |
| Visibility control | Hide pages that aren't ready, are in internal testing, or should only be visible to specific roles |
| Ordering | Reorder menus and pages so the runtime system matches how business users work |
Menu management answers the question "how do business users find and use the system?" Once AI generates pages, menu management turns them into a coherent, usable business system.
Component-level page permissions
Component-level permissions control what a user can do inside a page — not just whether the page is visible.
| Permission target | Examples |
|---|---|
| Buttons | Create, edit, delete, export, bulk update, approve, submit, and so on |
| Fields | Hide sensitive fields, make them read-only or editable, or show them per role |
| Table actions | Row actions, bulk actions, filter conditions, export |
| Page sections | Restrict statistic cards, charts, or detail blocks to specific roles |
| Agent actions | Restrict whether the Agent may write, delete, or export on the user's behalf |
Component-level permissions match real enterprise scenarios: on the same customer list, sales, supervisors, finance, and operations may each see completely different fields and available actions.
Advanced page editing
Advanced page editing is for fine-tuning pages generated from AI or a database.
| Editing target | Description |
|---|---|
| Layout | Adjust page structure and the arrangement of table areas, filter bars, detail areas, and action areas |
| Field configuration | Control display names, ordering, visibility, width, format, and enum rendering |
| Component configuration | Configure tables, forms, charts, buttons, dialogs, tags, status components, and more |
| Interaction configuration | Configure navigation, dialogs, linked updates, validation, defaults, and bulk actions |
| Data binding | Adjust the datasets, fields, query conditions, and action entries a page is bound to |
Advanced page editing lets product, implementation, and engineering teams keep polishing generated pages until they meet the usability and detail standards of a production system.
Conversational editing with the Agent
Conversational editing lets users change pages in natural language instead of grappling with configuration options.
| What you say | What happens |
|---|---|
| "Put customer level after customer name" | Table field order is adjusted |
| "Add a filter so I only see Grade-A customers" | A filter field with a default condition is added |
| "Add a deal-amount trend chart to this page" | A dashboard block or chart component is generated |
| "Hide the delete button for regular users" | Component-level permissions are configured |
| "Generate an analysis page from the sales funnel" | A sales funnel dashboard is generated |
This works well for product, operations, and business staff joining page iteration — and for developers who want to hand page tweaks to the Agent.
Dashboard generation
Dashboard generation turns enterprise data into visual analytics pages.
| Analysis type | Description |
|---|---|
| Metric cards | Show headline numbers: total leads, opportunity value, order revenue, payments received, conversion rate |
| Trend analysis | Track new additions, closed deals, payments, inventory, and expenses over time |
| Funnel analysis | Follow conversion from lead to opportunity, quote, order, and signed contract |
| Distribution analysis | Break data down by region, channel, customer grade, product type, and more |
| Comparison analysis | Compare performance across teams, stores, cities, channels, and products |
Dashboards can be generated from a database structure, custom SQL, a dataset, or a natural-language request. Managers, operations staff, and business owners can use it to get an analytical view quickly.
Page development covers both standard management pages and fast generation of analytics pages, operations boards, and ad-hoc management views.
3. RabetBase
RabetBase is Lovrabet's data and backend foundation. It connects enterprise databases, understands the business data structure, accumulates reusable data capabilities, and turns raw database power into business capabilities that pages, APIs, Agents, and external tools can all call.
Business data domain model
The business data domain model lifts an enterprise database from "tables and columns" to "business objects and relationships".
| Capability | Description |
|---|---|
| ER diagrams | Visualize relationships between tables, showing how customers, orders, products, contracts, employees, and other objects connect |
| Business object recognition | Identify tables as business objects — customers, contacts, opportunities, orders, payment records |
| Field semantics | Understand what each field means — its type, enum values, primary and foreign keys, and business purpose |
| Relationship analysis | Analyze one-to-one, one-to-many, and many-to-many relationships to support page generation and Agent queries |
| API and page mapping | Determine which APIs, list pages, detail pages, create pages, and edit pages each object can produce |
The domain model is the prerequisite for everything Lovrabet generates. Only by first understanding the objects and relationships in your enterprise database can pages, datasets, smart search, and Agent execution stand on solid ground.
Database connector
The database connector links Lovrabet to your existing databases and analyzes them intelligently.
| Capability | Description |
|---|---|
| Database binding | Connect existing MySQL or other databases as data sources for app building |
| Connection testing | Verify database connectivity and configuration |
| Intelligent analysis | Analyze tables, fields, and relationships to produce the business data domain model |
| Incremental analysis | Re-analyze after database schema changes and update the model's understanding |
| Secure access | Control the accessible database scope via connection settings and permission policies |
The connector's value: Lovrabet never starts from a blank slate — it starts from the business data and system assets the enterprise already has.
Custom SQL management
Custom SQL management captures the metric definitions, complex queries, and high-value analysis logic your enterprise relies on.
| Capability | Description |
|---|---|
| Create and maintain | Create, edit, test, and manage custom SQL |
| Unified metric definitions | Capture sales funnels, acquisition quality, payment amounts, order revenue, and other metrics as shared queries |
| Page reuse | Dashboards, list pages, and analysis pages all reuse the same SQL definitions |
| Agent calls | The Agent can run SQL queries to answer business questions in natural language |
| History | Keep SQL change and execution records for traceability and governance |
Custom SQL fits complex statistics, cross-table analysis, and KPI calculations that plain table queries can't express — so you never implement the same metric twice across pages, reports, and Agents.
Backend Function
Backend Functions inject custom business logic before and after standard data operations — the backend extension point for complex business rules.
| Function type | Description |
|---|---|
| Before functions | Run before a data write — permission checks, parameter validation, field completion, business rule validation |
| After functions | Run after a data write — data synchronization, status updates, notifications, audit records |
| Endpoint functions | Expose a standalone HTTP endpoint for complex server-side logic or external system calls |
Backend Functions handle logic that standard CRUD can't: data permission isolation, password salting, sensitive-field masking, cross-table validation, inventory deduction, contract status changes, or updating upstream records after an order is signed.
Open API
Open API exposes Lovrabet app capabilities to external systems and apps on any platform.
| Capability | Description |
|---|---|
| RESTful APIs | Standard interfaces for querying, creating, updating, deleting, aggregating, and exporting data |
| Authentication | Control external access identities with AccessKey and similar mechanisms |
| Cross-platform calls | Call from internal systems, mini-programs, mobile, desktop, and third-party platforms |
| SDKs | Call app data and business capabilities through SDKs |
| API documentation | Generated interface descriptions, parameter structures, and field references for every business object |
With Open API, a Lovrabet-generated app is more than a page system — it becomes a capability endpoint for internal services and external ecosystems.
Dataset management
Dataset management packages database tables, custom SQL, Open APIs, and Backend Functions into reusable data capabilities for pages and Agents.
| Capability | Description |
|---|---|
| Source packaging | Wrap database tables, SQL, and APIs into uniform datasets |
| Field management | Manage field names, types, display formats, enum values, and business meanings |
| Query capabilities | List queries, detail queries, filtering, aggregation, export, and option lookups |
| Page generation | Datasets directly drive list, detail, create, and edit pages, plus dashboards |
| Agent calls | Agents query, analyze, and execute business actions on datasets in natural language |
Datasets are the middle layer connecting databases, pages, APIs, and Agents. Data capabilities get reused across many entry points instead of being locked to one page or one piece of code.
RabetBase's core value is understanding the enterprise's existing databases and business flows before generating management modules. For temporary statistics and management needs, text-to-system generation gets you there fast; for highly customized, interaction-heavy, or deeply integrated capabilities, keep building with the RabetBase CLI.
2. Lovrabet runtime app system
The runtime app system is where business users actually work. The conversational Agent is the default first entry point, while classic standard system pages remain available — switch between them from the top-left corner.
1. Conversational Agent system
The conversational Agent system lets business users state a goal in natural language; the system automatically invokes pages, data, SQL, Skills, BFFs, and external tools to get it done.
| Capability | Description |
|---|---|
| Natural-language entry point | Describe the goal directly — query data, submit a request, fill in records, generate statistics, or check business trends |
| Skill invocation | The Agent calls configured enterprise Skills — lead-to-opportunity conversion, lifecycle state changes, updating upstream records after a signed order, and more |
| Data and action execution | Using datasets, custom SQL, Backend Functions, and similar capabilities, the Agent queries, analyzes, writes, and coordinates |
| Business context awareness | The Agent interprets intent from the current app, current page, user permissions, and business data |
| Task output and review | Produces analysis conclusions, business suggestions, operation records, and next actions |
2. Classic standard system pages
Alongside the conversational Agent, the runtime app keeps classic standard system pages for users who want a stable interface, bulk operations, structured form entry, and structured management.
| Page type | Description |
|---|---|
| List pages | Show business object lists with filtering, sorting, pagination, bulk actions, and links to details |
| Detail pages | Show a record's complete information, related data, and available actions |
| Create / edit pages | Add or modify business data, with field validation, permission control, and backend logic execution |
| Dashboards | Show business metrics, charts, trends, funnels, distributions, and retrospective analysis |
| Menus and navigation | Organize customer, sales, product, transaction, and organization modules by business domain |
3. AI-powered search
List filtering in the runtime system supports AI-powered search. No need to decode complex field names and filter conditions — just express what you want in natural language.
| Use case | Example |
|---|---|
| Conditional queries | "Find Grade-A customers added in the East China region in the last 30 days" |
| Business filters | "Show opportunities with an expected deal value over 50,000 CNY that aren't signed yet" |
| Anomaly hunting | "Find customers with high amounts that haven't been followed up in a long time" |
| Combined queries | "Show customers created this month in the Delta region with no owner and Grade B" |
AI-powered search converts natural language into executable filter conditions and returns matching data within the user's permissions.
3. CLI capabilities
The CLIs are Lovrabet's key entry points for Agent tools, enterprise systems, and developer workflows.
1. Lovrabet CLI
The Lovrabet CLI lets any Agent tool interact directly with the Lovrabet runtime app system. It connects enterprise capabilities — Feishu/Lark, enterprise IM, workflow systems, knowledge bases, automation tools — into the business system.
| Capability | Description |
|---|---|
| Agent tool integration | External Agents query, create, update, and analyze business data in Lovrabet apps |
| Enterprise system coordination | Work with Feishu and other enterprise tools for messaging, documents, approvals, tasks, and business system coordination |
| Runtime data interaction | Call runtime datasets, SQL, APIs, and business actions directly |
| Automation orchestration | Wrap system operations into Agent-callable tools for cross-system automation |
2. RabetBase CLI
The RabetBase CLI targets developers and intelligent development tools, supporting app development, data modeling, page generation, BFF writing, and capability publishing inside AI development environments such as Codex.
| Capability | Description |
|---|---|
| Intelligent tool integration | Drive RabetBase capabilities directly from tools like Codex for AI-assisted development |
| Database and model development | Manage database connections, tables, fields, relationships, and datasets |
| Page and module generation | Generate pages, menus, and module structures from business objects |
| Custom SQL and BFF development | Write and maintain complex queries, metric definitions, and backend business logic |
| Deep custom development | Extend with enterprise-specific capabilities where standard generation falls short |
4. How the capabilities combine
Lovrabet's basic features aren't isolated modules — they form a closed loop around building and running enterprise apps.
| Stage | Recommended combination | Output |
|---|---|---|
| Understand the business | Database connector, business data domain model, dataset management | Identify enterprise databases, ER relationships, and business objects |
| Generate the system | Page development, menu management, dashboard generation, text-to-system generation | Produce management pages, analytics pages, and business modules |
| Accumulate capabilities | Custom SQL, Backend Functions, Open API, Skills, Knowledges | Build reusable data, action, API, and AI assets |
| Run and use | Conversational Agent, classic pages, AI-powered search | Support natural-language operations and standard page management for business users |
| Govern securely | App permissions, page component permissions, roles and members, notification configuration, audit logs | Ensure enterprise-grade permissions, security, and collaboration governance |
| Open up and extend | Lovrabet CLI, RabetBase CLI | Support enterprise system integration and intelligent development tooling |
In a nutshell
Lovrabet's basic product features span three layers: the admin console organizes enterprise databases, business logic, pages, AI assets, and permission governance; the runtime lets business users get work done through the Agent and standard pages; and the CLIs let external enterprise systems and intelligent development tools take part directly in app building and business execution.