> ## Documentation Index
> Fetch the complete documentation index at: https://docs.snapsec.co/llms.txt
> Use this file to discover all available pages before exploring further.

# Getting Started with AI Pentesting

> Launch your first autonomous penetration test in four simple steps.

Follow this guide to configure and launch an AI-driven penetration test.

***

## Step 1: Define Your Assessment Scope

1. Navigate to **AI Pentesting > Projects & Assessments**.
2. Click **New Assessment**.
3. Provide an assessment title (e.g., `Q4 Web Application Pentest`).
4. Specify target root domains, API base URLs, or IP ranges.
5. Define explicit out-of-scope endpoints (e.g., `/auth/delete-account`).

***

## Step 2: Configure Rules of Engagement & Safety

1. Select test intensity (Passive, Standard, or Aggressive).
2. Set rate limits (e.g., maximum 20 requests per second).
3. Choose **Autonomous Mode** (fully automated) or **Human-in-the-Loop Mode** (requires confirmation before executing high-impact exploits).

***

## Step 3: Provide Test Credentials

Provide credentials for two different test user accounts to enable multi-tenant authorization and privilege escalation testing:

* **Test Account 1 (Standard User)**: `user1@test.com`
* **Test Account 2 (Admin / Peer User)**: `user2@test.com`

***

## Step 4: Execute & Review Live Findings

1. Click **Start Assessment**.
2. Watch the AI agent stream its decision-making log, crawled attack tree, and validated findings in real time.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.