Scope and rules of engagement
We agree in writing what to test, when and within which limits. We only work on authorised systems.
We find out what is exploitable before someone else does, with evidence for NIS2 and ISO/IEC 27001.

A scan can return hundreds of vulnerabilities. The useful work is understanding which ones can be exploited, what they give access to and what they would cost the organisation. We combine vulnerability assessment and penetration testing on the external perimeter, the internal network, web applications, APIs and cloud configurations. Applications with LLMs and AI agents have a dedicated track: AI security.
For web applications and APIs we use the OWASP Top 10 and the OWASP API Security Top 10 as references.
We agree in writing what to test, when and within which limits. We only work on authorised systems.
We map domains, exposed services, applications and technologies, the way an attacker would.
Systematic scans of networks, web apps and APIs, with the results checked to remove false positives.
Controlled exploitation of vulnerabilities to understand what can be reached and with what impact.
For every risk: evidence, business impact and a proposed fix. Priorities take into account CVSS, EPSS and the CISA KEV catalogue.
After the fixes we check that the vulnerabilities are closed and update the evidence.
The attack surface changes every week, while a penetration test captures a single moment. CyberAgent keeps networks, web apps and APIs under control with continuous vulnerability assessment and automated pentesting. Before every scan it verifies domain control, so it only works on the customer's own assets. Risks arrive validated and ranked by impact, with evidence for NIS2 and ISO/IEC 27001; the decision on what to fix stays with the team.
CyberAgent also runs simulated phishing campaigns for staff, with results aggregated by department.
The obligations on vulnerability management are covered on the NIS2 page.
A vulnerability assessment systematically identifies the known vulnerabilities of systems and applications. A penetration test simulates a controlled attack to check which ones can be exploited and how far an attacker can get.
We recommend at least once a year and after every significant change, such as a new application or a migration. Between tests, continuous vulnerability assessment keeps the attack surface under control.
We agree on time windows, limits and emergency contacts before starting. Tests that may degrade a service are only run if authorised and at agreed times.
Yes. The report documents vulnerabilities, priorities and fixes and is evidence for the vulnerability management required by NIS2 (measure ID.RA-08) and by control 8.8 of ISO/IEC 27001:2022.
Fill in the form: your request goes to the dedicated team.