Overview

Quick start

From a fresh server to the report of your first scan: install, sign in, connect AI, create a product, scan.

View as Markdown

From a fresh Linux server to the report of your first scan, in six steps. Each step says how to check it worked and links to the page with every option.

Prefer to hand this to an AI agent?

A coding agent with a shell on the server can run the install (step 1) for you and stop at the steps only you can do. Give it this page: https://scansuite.eu/documentation/quick-start.md. Once the server runs, Connect your AI agent lets it scan for you too.

Before you start

You needDetailsCheck
A Linux serverUbuntu 22.04 or newer recommended, a non-root account with sudo. Sizes and other systems in Technical Requirements.. /etc/os-release; echo $ID $VERSION_ID
The licence fileFrom ScanSuite: <name>_<LICENCE_CODE>.licls *.lic in the folder you install from
The registry tokenFrom ScanSuite, starts with dckr_pat_You type it when the installer asks
An AI modelAn OpenAI-compatible endpoint (hosted or on your network) or Vertex AI (Claude), with its API keyNeeded in step 3
A repository to scanAny Git repository you own; a public HTTPS URL needs no key—

Full requirements: Technical Requirements.

1. Install

In a folder that holds exactly one licence file, run:

Shell
bash -c "$(curl -sSL https://scansuite.eu/install-teams)"
A person does this step

When the installer prints "Signing in to the image registry" and asks for a Password, type the registry token. If the run had no terminal (for example an agent's shell), sign in once yourself with sudo docker login -u <REGISTRY_USER> --password-stdin and run the installer again: running it twice is safe.

Check: the output ends with [+] All services healthy and [+] ScanSuite Teams is installed in /opt/scansuite. Then:

Shell
cd /opt/scansuite && ./scansuite doctor

It ends with [+] No problems found. If not, see Troubleshooting. Every installer option is in Install ScanSuite Teams; for Google Cloud or Azure use Install on Google Cloud or Install on Microsoft Azure instead.

2. Create the first account

A person does this step

Open https://<YOUR_HOST>/setup in a browser (accept the self-signed certificate warning once, or set up a certificate first: SSL/TLS Setup). Fill in Team name, User name, Password and Confirm password and press Create the first account. The password needs at least 8 characters with upper and lower case, a digit and a special character.

Do this straight after installing: whoever opens the page first gets the account.

Check: you land on the sign-in page and can sign in. This account is team admin of the first team and installation administrator; /setup is gone for good.

3. Connect an AI model

The scan forms start with the AI scanners selected, and they need a model. As the installation administrator, open the user menu (top right) → System Settings → Shared services:

  1. AI connections → Add connection: a name, the kind (OpenAI-compatible or Vertex AI (Claude)), the API URL, and the API key → Save.
  2. AI models → Add model: a name, the connection, the model ID → Save. The first model answers every AI task.
A person does this step

Type the API key (or the Vertex service-account JSON) into the form yourself. ScanSuite may ask you to sign in again before it saves AI settings.

Check: the model shows Usable, and System AI → Test connection says the system AI answered. Every team uses the system AI unless it sets its own. Details, per-task routing and keeping code on your network: AI providers and cost.

4. Create a product

A product groups everything scanned for one application. In the sidebar open Products → New product, enter a Product name (unique in the team) and, optionally, the Repository URL → Create product.

Check: the message Product created or updated, and the product in the list. More, including creating products from the API: Products.

5. Run your first scan

In the sidebar open Static Code Analysis → Git Repository:

  1. Repositories: the clone URL, for example https://github.com/<org>/<repo>. (Or use Upload ZIP for an archive of the code.)
  2. Product Assignment: the product from step 4.
  3. Leave the rest at its defaults: Full scan, all supported languages, and the AI scanners (AI Native SAST, secrets with AI verification, dependency check with reachability).
  4. Press Start Analysis.

Check: Scan submitted, and the scan page opens and shows progress. A private repository needs an SSH key first: Git repository authentication. If the scan log says AI SAST cannot start, step 3 is not done.

6. Read the results

  • The report: when the scan finishes, Download report on the scan page (or Report on its row in Scan History) gives a ZIP with HTML, XLSX and JSON.
  • The findings: sidebar Vulnerabilities, with the product selected. Each finding carries its validation verdict and the evidence behind it: reading a finding.
  • Secrets found in the code: sidebar Credentials.

Next

Web App Scanning

Scan websites and web APIs, signed in or not, with classic scanners, AI DAST or the AI Pentest agent; the target rules for every scan.

Infrastructure Checks

Vulnerability scans, network discovery, patching checks, domain OSINT and container images, with the AI Pentest agent as the default.

Teams and roles

What each of the four roles can do, and how a team admin sets up a team: targets, AI, scan credentials, reports and API tokens.

CI/CD and automation

Run SAST, DAST and infrastructure scans from GitLab, GitHub, Jenkins, Azure DevOps, Bitbucket or a script with a team API token, test the token with --check, and fail the build on what they find.

Connect your AI agent

One line for Claude Code, Codex, Copilot or Cursor: the agent installs the ScanSuite skill and connects to your server.

Schedule the scan

Schedule web, infrastructure and static scans, manage them on the Schedules page, or schedule from your CI system.

Web and infrastructure scans only run against targets the team allows; a team admin sets them on the team page (Teams and roles).

Last reviewed 2026-10-19