gateapi-firewall

---
name: gateapi-firewall
description: Enterprise security firewall and authorization control plane for OpenClaw agents
version: 1.0.0
author: GateAPI
tags: [security, guardrails, authorization, enterprise, firewall]
category: Security
website: https://agent-gate-rho.vercel.app
---

# GateAPI - Enterprise Security Firewall for OpenClaw

Stop your OpenClaw agent from executing dangerous commands, spending money, or taking irreversible actions without your explicit approval.

## The Problem

By default, OpenClaw has unrestricted access to your system. One hallucination can:
- Delete your files with rm -rf
- Send emails to the wrong people
- Make unauthorized API calls or payments
- Push broken code to production

GateAPI puts YOU back in control.

## How It Works

GateAPI intercepts every tool call your OpenClaw agent makes BEFORE it executes. It checks the action against your human-defined policy rules and returns one of three decisions:
- ALLOW - action executes normally
- DENY - action is blocked, agent is warned
- REQUIRE_APPROVAL - action pauses, you get a Telegram/email notification to approve or reject

## Features

- Block dangerous bash commands (rm -rf, drop table, etc.)
- Require human approval for high-risk actions via Telegram or email
- Real-time audit log of every action your agent attempts
- Visual dashboard to manage all your agent security rules
- AI Policy Wizard - describe rules in plain English, auto-generates the regex
- Works with any OpenClaw agent in 2 lines of code

## Setup (2 minutes)

1. Go to https://agent-gate-rho.vercel.app and create a free account
2. Create an Agent profile and copy your API key
3. Wrap your OpenClaw agent with the GateAPI SDK

## Installation

npm install @gateapi/openclaw-guard

## Usage

import { withGateAPI } from '@gateapi/openclaw-guard';

const securedAgent = withGateAPI(myAgent, {
  apiKey: "your_api_key_from_dashboard",
  endpoint: process.env.GATEAPI_ENDPOINT || "https://agent-gate-rho.vercel.app"
});

await securedAgent.execute("Start task");

## Example Policies

- Block all rm -rf bash commands -> DENY
- Require approval for any payment over $50 -> REQUIRE_APPROVAL
- Block access to any URL outside your whitelist -> DENY
- Require approval before sending any email -> REQUIRE_APPROVAL

## Dashboard

Manage all your agent rules, view real-time audit logs, and set up approval webhooks at:
https://agent-gate-rho.vercel.app

## Pricing

Free for MVP. No credit card required.