Why Cyber Baseline
Useful knowledge.
Safer systems.
Security knowledge should help people make better decisions and build safer systems. Cyber Baseline exists to make complex Microsoft security guidance clearer and easier to apply.

Behind the publication
Fouad Mulla
Created by Fouad Mulla, Microsoft Certified Trainer (MCT), Cyber Baseline brings together practical guides, detection queries and enterprise architectures to support defenders, builders and learners.
A baseline you can measure against
Define the identities, access and behavior you expect, then gather the evidence to test them. Our practice loop is Understand · Constrain · Observe · Respond: map the system, limit authority, collect useful signals and use investigations to improve the next baseline.
Start with the security boundary
A model is one part of a system. Identity, retrieval, connectors, application code and operating practices all shape what that system can do. Our guides begin with those boundaries and the evidence needed to test them.
Make the review status clear
Every architecture, guide, query recipe and campaign insight shows its sources, review date and reviewer. Documentation checked means the claims were checked against linked primary sources, including Microsoft documentation and original threat research. AI reviews are attributed to Codex (AI). This review covers product scope, documented controls and limitations; it does not claim independent human review or testing in a live tenant. A technical review is recorded separately when it happens.
Use AI to draft, people to decide
AI-assisted content begins as an unpublished draft. Claims and examples are checked before publication, and a maintainer decides whether to release the result. Automated drafting cannot approve or publish its own output. Unpublished drafts are excluded from the site, search, feeds and sitemap.
Explain with diagrams, test with evidence
Our security diagrams use original layouts created for this publication; enterprise architecture views use official Microsoft service icons to identify components. They show the relationships between controls, not Microsoft product screenshots or evidence that a deployment has been tested.
Make architecture decisions testable
Enterprise reference architectures connect a specific use case to resource boundaries, numbered flows, service choices and operating responsibilities. Each design records its assumptions, relevant SKU or regional constraints, rollout sequence, monitoring and recovery requirements, and acceptance tests. The accompanying diagram shows the intended design; teams must validate routing, authorization, telemetry and recovery in their own environment before adopting its operating commitments.
Keep intelligence dates and scope explicit
Campaign insights focus on one issue, campaign or CVE. Report dates describe the cited evidence; review dates describe our latest source check. Historical exploitation is not a claim of activity today. We distinguish affected products from defensive tools, vendor fixes from additional hardening, and contextual queries from validated detection rules.
Keep the guidance grounded
We prefer Microsoft documentation and other primary sources. A recommended control is not a guarantee, and availability may depend on product, deployment mode, region or licensing. Check the source, test expected and denied behavior, and capture the result.