iDox.ai Guardrail™

Real-Time AI Agent Security & Data Protection

Stop sensitive data from being pasted or uploaded into AI tools like ChatGPT. Block risky prompts. Control AI agents like OpenClaw – and keep AI use secure with AI agent security controls.

Install iDox.ai Guardrail™

Run a single command to install and start using iDox.ai Guardrail™ instantly.

Currently supports Windows only.

iwr -useb https://idox.ai/install-guardrail.ps1 | iex

What it does

Prevent data leaks before text or files reach AI tools

Prevent data leaks before text or files reach ChatGPT or other AI tools

Detect risky prompts and unsafe instructions

Detect risky prompts (prompt injection, unsafe instructions)

Control AI agent access with allow or deny prompts

Control AI agent access (including OpenClaw) with allow/deny prompts

Choose your edition

Individual

Real-Time Guardrail for Pros & Small Teams

On-Device Interception: Block sensitive data before it reaches ChatGPT or other AI tools.
Agent Communication Monitoring: Audit data flow between AI agents (like OpenClaw) to prevent leaks.
Abnormal Command Detection: Identify and block prompt injections or unauthorized agent actions.

Explore Individual

Enterprise

Centralized Governance for Enterprise AI

Unified Policy Engine: Enforce "Block or Sanitize" rules for all employee AI usage.
Full Audit & Abnormal Command Tracking: Monitor high-risk Agent behaviors with "who-what-when" visibility.
Hosting & On-Premises Options: Flexible deployment (Cloud or Local) to meet strict compliance needs.

Explore Enterprise

How it works

Detect

Intercept

Sanitize or Allow

(in the moment, on-device)