Skip to content
HuntSecAI
Hunt Web Documentation

Introduction

Hunt Web is an AI-assisted web security testing tool that combines a CLI and a browser-based Dashboard.

Hunt Web is an AI-assisted web security testing tool that brings a command-line scanner and a browser-based Dashboard into one product. It helps users define targets, run assessments, inspect findings, and continue testing through Hermes Chat. Hunt Web connects these steps around saved scan context, so follow-up work can build on requests, evidence, and results from an earlier scan.

Figure 1. The Hunt Web logo.
Hunt Web logo with silver HW lettering on a red and dark circular emblem

Who Is Hunt Web For?

Hunt Web is designed for pentesters, security researchers, and development teams assessing applications they own or are authorized to test. It supports the practical work of preparing an assessment, exploring application behavior, following up on potential issues, and reviewing the evidence behind findings.

A pentester can use it to organize separate targets and scan sessions. A researcher can return to saved requests and direct the next investigation. A development team can review findings and reports when deciding what to investigate or fix. The Dashboard makes this work visible through targets, scan activity, severity filters, and reports.

From CLI to an Integrated CLI and Dashboard

Hunt Web began as a CLI tool and now includes a Dashboard alongside the CLI and the shared scanning runtime. Users can work through terminal commands or use the Dashboard to prepare targets, configure access, monitor scans, and inspect results.

The Dashboard overview brings together active targets, running and paused scans, recent Runs, and verified findings grouped by severity. This gives users a place to see current work and choose what to examine next.

Figure 2. The Dashboard provides an overview of scan activity and saved findings.
Hunt Web Dashboard showing scan metrics, Recent Runs, and Findings by severity

Testing Modes and Core Capabilities

Hunt Web supports three approaches to application testing:

Blackbox

Explore the application without supplying authenticated accounts.

Greybox Environment

Test through HTTP requests using configured account access.

Greybox Browser

Test through an interactive Chromium session with saved account access on a Linux Desktop.

Greybox workflows support one or two accounts, allowing checks involving different users or roles when those accounts are configured. Scan profiles include Quick, Standard, and Deep. Users can monitor a Run, pause it, and resume work using its saved context.

Targets and Scope define the application and the boundaries of an assessment. Runs hold its execution history. Findings provide access to verified issues and supporting evidence, while Reports organize saved assessment results for review and PDF export.

The combination is useful when an investigation requires several passes: users can inspect what happened, identify a question worth pursuing, and continue from the same assessment context.

How Hermes Fits Into the Workflow

Hermes Chat connects conversation with the selected Run. It can read saved scan results, request history, and earlier work across Blackbox, Greybox Environment, and Greybox Browser.

Users can ask about previous activity or give instructions for further testing. For example, a question about the current target can be answered from saved context, while a request to check whether a Browser account is still logged in requires a fresh access check. An instruction to continue testing can launch further scanner work against the Run's saved Scope.

This keeps the conversation connected to the assessment. Chat Templates also provide starting points for directing follow-up investigations.

A Typical Assessment

The main workflow follows Target → Scope → Run → Findings → Report:

  1. 1TargetAdd the application and choose the testing mode.
  2. 2ScopeSave the URLs and account access to test.
  3. 3RunPick a profile and follow the scan.
  4. 4FindingsReview verified issues and evidence.
  5. 5ReportRead the saved snapshot or export a PDF.
Conceptual overview of the workflow.
  1. Create a Target. Add the application and choose the testing mode.
  2. Prepare the Scope. Define the application URLs and configure account access where needed. Check authentication before starting a Greybox scan.
  3. Start a Run. Select a scan profile and follow its activity and requests.
  4. Review Findings. Examine verified issues and their evidence, then use Hermes Chat to investigate further questions.
  5. Read or export the Report. Review the saved assessment and download a PDF for sharing or further analysis.

Snapshots and human review

Reports use saved assessment snapshots, so they can lag behind new findings during an active scan. Automated verification supports the review process; human review remains necessary when assessing the conclusions.

Getting Started

Hunt Web uses one installer with Linux Desktop and Linux Server/VPS options. The tested platform is Ubuntu 24.04 on x86_64. Desktop supports all three testing modes, including interactive Chromium. Server supports Blackbox and Greybox Environment, with headless Chromium available for tools that need it.

Start with the Linux Desktop or Linux Server/VPS installation guide included in the package. After installation, configure the model provider and account access, sign in to the Dashboard, and prepare the first Target. For a VPS, the Server guide also explains the separate domain and HTTPS setup needed to make the Dashboard accessible remotely.