SupraWall Documentation

The open standard for AI agent security. Learn how to secure, monitor, and govern your agents in less than 5 minutes.

5-Minute Quickstart

Get your API key and secure your first agent. The fastest way from zero to secured.

Get started

Policy Spec (AGPS)

Read the open standard specification for AI tool permissions and policy formatting.

Read the standard

Interactive Framework Guides

LangChainLlamaIndexAutoGenVercel AI SDKCrewAI

Core Concepts

Allow Policies

Explicitly allow safe tools like reading files or searching the web without human intervention.

Deny Policies

Hard-block destructive operations like deleting directories or calling restricted APIs.

Human Approval

Pause execution via Webhook or Dashboard when a tool attempts to send emails or drop tables.

Real-time Audit

Every single decision is logged securely in the cloud to provide an instant trail of AI actions.