AgentWall Documentation

AgentWall Documentation

Guard the Agent, Save the Budget 🛡️

AgentWall is the world's first "Agent Firewall" for AI Agents. It provides run-level governance, loop detection, budget enforcement, and DLP protection for your AI agents.

Section Description
Getting Started 5-minute quickstart guide
SDK Examples Python & JavaScript code examples
API Reference OpenAI-compatible API docs
Concepts Core concepts and architecture
Playground Interactive examples

Why AgentWall?

The Problem

AI Agents are powerful but dangerous:

  • Runaway costs: An agent can spend $50,000 overnight in an infinite loop
  • Data leaks: Sensitive data (API keys, PII) can leak through prompts
  • No visibility: You don't know what your agents are doing until it's too late

The Solution

AgentWall sits between your agents and LLM providers:

Key Features:

  • 🔄 Loop Detection - Stops infinite loops at the 2nd request
  • 💰 Budget Enforcement - Per-run, daily, monthly limits
  • 🔒 DLP Protection - Masks API keys, credit cards, PII
  • 📊 Run Tracking - See every step of your agent's journey
  • <10ms Overhead - Ultra-low latency proxy

Quick Example

Production Ready

AgentWall is battle-tested and production-ready:

Metric Value
Uptime 99.9%
Latency Overhead <10ms
Loop Detection 100% accuracy
DLP Patterns 15+ types

Get Started

  1. Sign up for an API key
  2. Follow the Getting Started Guide
  3. Explore the API Reference

Need help? Contact us or check our FAQ