Documentation - Guardrail Layer

Documentation

Technical documentation for Guardrail Layer — covering architecture, enforcement models, security boundaries, and real-world deployment patterns.

Getting Started

Overview of Guardrail Layer, core concepts, and how it fits between LLMs and production databases.

View

Metadata

How Guardrail Layer uses schema metadata to understand tables, columns, relationships, and sensitivity boundaries.

View

Roles & Users

Role-aware enforcement, permission models, and how access decisions are evaluated per user and agent.

View

Query Enforcement

How queries are validated, rewritten, allowed, or denied before execution.

View

API Token Creation

Guide that shows you how to create an API token in Guardrail Layer

View

Audit Logging

Complete execution trails for every query, including decision context, rewrites, and enforcement outcomes.

View

Deployment & Architecture

Guardrail Layer can be deployed as a managed service or fully self-hosted within your own infrastructure. The enforcement model remains identical across both approaches.

These documents reflect real production constraints encountered while allowing LLMs to interact with live databases.
Scroll to Top