Learn how Shield enforces security rules at the AIX and Semantic layer to control who can access which data in TADA.
Table of Contents
Prerequisites
- Access to TADA platform tools with a Super User role
- A published architecture with security definitions in place
- User roles and permissions configured in the Admin Portal
What is Shield?
Shield is TADA's security enforcement layer, operating at the AIX & Semantic Security layer alongside the Configurator. Where the Configurator defines what the AI and the platform understand about your data, Shield controls which users can see and interact with which parts of it — enforcing access rules at the property, node, and data level.
With Shield, you:
- Define and enforce node-level security — restrict which roles can see a given entity
- Apply property-level security — hide or mask individual fields from specific roles
- Set row-level security — filter the data rows a user sees based on their role or context
- Configure security tollways — checkpoints the platform verifies before exposing data to the AI or end users
- Review the AIX security readiness — confirm that sensitive elements are properly blocked before publishing the intelligence layer
Shield Interface Overview
| # | Area | Description |
|---|---|---|
| 1 | Security Definitions | Access via Composer → Architect → select a node or link → Security tab. Define which roles have read, write, or no access to that entity. |
| 2 | Property-Level Rules | Within a node or link's Security tab, configure individual properties to be hidden, read-only, or fully accessible per role. Sensitive fields (e.g. salary, PII) can be masked entirely. |
| 3 | Row-Level Security | Scope the rows a user can see by binding security filters to their role or business element context. Users only see data that falls within their permitted scope. |
| 4 | Security Tollway | A platform checkpoint that verifies security rules are applied before data is served to the AI or end users. Flagged in the AIX Readiness Score when a node lacks required security definitions. |
| 5 | Role Management | User roles are created and assigned in the Admin Portal. Shield enforces those roles at runtime — every data request is checked against the caller's role before results are returned. |
| 6 | API Queue Security | Controls which roles can trigger API queue operations and architecture-level actions. Configured in Composer → Architect → Architecture Permissions. |
Key Concepts
| Term | Definition |
|---|---|
| Node-Level Security | Controls whether a role can access a given entity (node) at all — read, write, or none |
| Property-Level Security | Controls visibility and editability of individual fields within a node or link, per role |
| Row-Level Security | Filters the data rows returned to a user based on their role or assigned business element context |
| Security Tollway | A verification checkpoint in the AIX pipeline that ensures sensitive data is not exposed to the AI before security rules are confirmed |
| Role | A named permission profile assigned to users in the Admin Portal; Shield evaluates every data request against the caller's role |
| Block AI | A per-element toggle in the Configurator that excludes a node, link, or property from the AI layer entirely, regardless of Shield rules |
Related: Security Architecture & Definitions | Configurator — AIX & Semantic Intelligence Configuration | Architecture Permissions and API Queue Management | Company Settings — Admin Portal