Cybersecurity work spans four distinct activities: penetration testing of systems as they run, reviewing source code and architecture, assessing how the delivery process prevents defects, and carrying the risk decisions that follow. Each answers a different question, and combining them without intent produces a report nobody acts on.
Testing is scoped against public catalogues rather than a private methodology. Web and API assessments map findings to OWASP Top 10:2025 and the API Security Top 10; verification depth follows OWASP ASVS 5.0; adversary simulation references MITRE ATT&CK tactics. A finding therefore carries an identifier that auditors and other vendors already recognise.
Code and architecture review reach what runtime testing cannot: the design decision that makes a whole class of defect possible. That distinction matters when the same finding reappears across releases — at that point the fix belongs in the development process rather than in the individual ticket, and repeating the test will keep confirming the symptom.
Findings feed obligations that already apply. Evidence produced during a test supports ISO 27001 controls A.8.8 and A.8.29, the risk-management measures required by NIS 2 Article 21, and GDPR Article 32 — provided it is written to be reused rather than filed, which mostly means recording what was tested and when, not only what was found.