# AI Pentesting

**Escape AI Pentesting** is powered by **Cascade**, Escape's multi-agent pentest engine. Cascade explores your application like a human pentester, reasons about attack chains, and proves exploitability with concrete evidence. It combines large language model reasoning with Escape's crawling and DAST infrastructure to find complex, multi-step vulnerabilities.

This page explains what changed with Cascade and what's available today. To launch an assessment, follow the [Quickstart](<https://docs.escape.tech/documentation/ai-pentesting/quickstart/index.md>).

## What Changed: From Fixed Agents to One Engine

Earlier, AI Pentesting shipped as a set of separate, single-purpose agents. Each vulnerability class had its own hardcoded agent: an XSS Agent, a SQLi Agent, an IDOR/BOLA Agent, and so on. Each ran with a fixed playbook and a fixed slice of the application.

Cascade replaces that model with a single adaptive engine:

- **One orchestrator, many specialists.** Cascade plans the engagement, then spawns focused worker agents on demand: reconnaissance, targeted exploitation, validation. The mix of workers depends on what the application actually exposes, not on a fixed agent list.
- **Shared context.** Workers exchange findings and discoveries as they go, so a lead found by one worker (a database-backed endpoint, a leaked token, a tenant boundary) immediately informs the others.
- **Independent verification.** A dedicated reporter agent re-runs every candidate finding against the live target and collects its own evidence before filing an issue, which keeps the false-positive rate low.
- **Coverage tracking.** An advisory coverage agent watches which surfaces have been tested and pushes the orchestrator toward the gaps.

The vulnerability classes you knew as individual agents are now **capabilities of Cascade**. The pages under this section describe how Cascade tests for each class and how to steer it.

## What's Available Today

- **Black-box multi-agent pentesting.** Cascade runs an autonomous swarm against web applications, REST APIs, and GraphQL APIs. See [The Cascade Engine](<https://docs.escape.tech/documentation/ai-pentesting/multi-agent-pentest/index.md>).
- **Broad vulnerability coverage.** Identity and access control, business logic, LLM and AI-specific security, injection, SSRF and file inclusion, insecure data handling, browser security, data integrity, secrets and cryptography, protocol-specific checks, and platform or framework checks. See [How It Works](<https://docs.escape.tech/documentation/ai-pentesting/how-it-works/#what-cascade-tests-for>) for the full list.
- **Unitary regression testing.** Upload a previous pentest report and Cascade replays each reported vulnerability one at a time to confirm whether it's still reproducible. See [Regression Testing](<https://docs.escape.tech/documentation/ai-pentesting/regression-testing-agent/index.md>).
- **Issue retest.** The next assessment replays open findings on its own. To recheck selected issues without a full assessment, use [Issue Retest](<https://docs.escape.tech/documentation/platform/issue-retest/index.md>). Same flow on DAST profiles.
- **Whitebox pentesting.** Attach repository source code and Cascade reasons over it alongside the running application, then validates findings dynamically. See [Whitebox Agent](<https://docs.escape.tech/documentation/ai-pentesting/whitebox-agent/index.md>).
- **Artifacts and context.** Attach source code for whitebox context, PDF pentest reports for regression replay, and OpenAPI exports, documentation, or screenshots as additional assessment context.
- **Authentication.** Configure one or more users with natural-language sign-in instructions. Multiple users unlock authorization testing across roles and tenants. See [Authentication](<https://docs.escape.tech/documentation/platform/authentication/index.md>).
- **Memory across assessments.** Cascade carries application knowledge between assessments, so issue finding stays consistent from run to run instead of starting cold every time.
- **Evidence-rich issues.** Every finding ships with an [attack chain](<https://docs.escape.tech/documentation/ai-pentesting/graph-reasoning/index.md>), an impact assessment (CVSS score and vector), and a [Proof of Exploit](<https://docs.escape.tech/documentation/ai-pentesting/proof-of-exploit/index.md>): the request chain, screenshots, and reproduction steps. The assessment page includes Summary, Reasoning, Issues, Coverage, and Logs tabs.
- **Simple setup.** A dedicated creation form walks you through scope, authentication, optional tuning, and a pre-launch review. No manual agent selection.

## Positioning: AI Pentesting, DAST, and ASM

### When to Use AI Pentesting

AI Pentesting is the right tool when you need:

- **Deep, adaptive testing**: agents that reason about application behavior and adapt
- **Complex vulnerability discovery**: multi-step attacks, business logic flaws, and authorization issues
- **Autonomous exploration**: agents that navigate complex workflows and understand context
- **Proven exploitability**: findings backed by a reproducible attack chain

### When to Use DAST

[Business Logic Aware DAST](<https://docs.escape.tech/documentation/dast/index.md>) is better suited for:

- **Systematic, repeatable testing**: rule-based checks with predictable coverage
- **CI/CD integration**: fast, automated security testing in pipelines
- **Custom rule enforcement**: organization-specific policies and governance

### When to Use ASM

[ASM](<https://docs.escape.tech/documentation/asm/index.md>) focuses on:

- **Attack surface discovery**: finding all exposed assets and endpoints
- **Asset inventory**: maintaining a view of your attack surface
- **Continuous monitoring**: tracking changes to your external exposure

## Guardrails

AI Pentesting supports kill switches, strict scope, path-level restrictions, rate limits, and max duration. See **[Guardrails](<https://docs.escape.tech/documentation/ai-pentesting/guardrails/index.md>)** and **[Scope](<https://docs.escape.tech/documentation/ai-pentesting/scope/index.md>)**.

## Getting Started

### Profile Setup

Create AI Pentesting profiles from **AI Pentesting \> [Profiles](<https://app.escape.tech/pentest/profiles/>)** by clicking **New Profile**. The dedicated creation form walks you through scope, authentication, optional tuning, and the pre-launch review. It's separate from the generic DAST scan-profile wizard.

You can edit an existing pentest profile from the profile page. Editing uses the same form, but the target URLs are read-only after creation.

For automation, use `escape-cli profiles create-ai-pentest` or `POST /v3/profiles/ai-pentest`. See [CLI Profiles](<https://docs.escape.tech/documentation/automate/cli/profiles/#creating-ai-pentesting-profiles-beta>) and the [Public API](<https://docs.escape.tech/documentation/automate/public-api/#creating-ai-pentesting-profiles>). The form's **Schedule** setting also lets you choose a future start time.

Ready to run your first assessment? See the [Quickstart](<https://docs.escape.tech/documentation/ai-pentesting/quickstart/index.md>).

## Documentation

- **[Quickstart](<https://docs.escape.tech/documentation/ai-pentesting/quickstart/index.md>)**: Get started with your first AI Pentesting assessment
- **[Scope](<https://docs.escape.tech/documentation/ai-pentesting/scope/index.md>)**: Standard vs Strict, restrictions, and enforcement
- **[How It Works](<https://docs.escape.tech/documentation/ai-pentesting/how-it-works/index.md>)**: The Cascade workflow, from crawling to evidence
- **[The Cascade Engine](<https://docs.escape.tech/documentation/ai-pentesting/multi-agent-pentest/index.md>)**: Architecture, skills, and limits
- **[Graph Reasoning](<https://docs.escape.tech/documentation/ai-pentesting/graph-reasoning/index.md>)**: How Cascade builds attack paths across endpoints
- **[Proof of Exploit](<https://docs.escape.tech/documentation/ai-pentesting/proof-of-exploit/index.md>)**: Evidence, logs, and coverage
- **[Regression Testing Agent](<https://docs.escape.tech/documentation/ai-pentesting/regression-testing-agent/index.md>)**: Replay uploaded pentest findings to validate fixes
- **[Issue Retest](<https://docs.escape.tech/documentation/platform/issue-retest/index.md>)**: Automatic replay on the next assessment, or recheck selected issues without a full run
- **[Guardrails](<https://docs.escape.tech/documentation/ai-pentesting/guardrails/index.md>)**: Kill switch, rate limit, and max duration
- **[Whitebox Agent](<https://docs.escape.tech/documentation/ai-pentesting/whitebox-agent/index.md>)**: Source-aware testing when you attach repository code
- **[FAQ](<https://docs.escape.tech/documentation/ai-pentesting/faq/index.md>)**: Context, evidence, models, scope, and AI data handling

## Related Documentation

- [Business Logic Aware DAST](<https://docs.escape.tech/documentation/dast/index.md>): Traditional rule-based security testing
- [ASM](<https://docs.escape.tech/documentation/asm/index.md>): Attack surface discovery and management
- [Authentication](<https://docs.escape.tech/documentation/platform/authentication/index.md>): Set up authentication for assessments
- [Private Locations](<https://docs.escape.tech/documentation/platform/private-location/index.md>): Test internal applications
- [API Testing Configuration](<https://docs.escape.tech/documentation/dast/api-testing/index.md>): API testing configuration options
- [Frontend DAST Configuration](<https://docs.escape.tech/documentation/dast/webapp-testing/index.md>): WebApp testing configuration options