Lesson 15 — Cloud Security Explorer
Cloud risk is rarely caused by one isolated setting.
This lesson explains how Microsoft Defender for Cloud Security Explorer queries cloud assets and security relationships to investigate exposure, trace risky combinations and identify the resources that matter most.

What you will learn
This lesson explains how to explore cloud assets, relationships and exposure paths.
Cloud Security Explorer investigation flow
Common query building blocks
| Query element | What it examines | Example investigation |
|---|---|---|
| Resource type | Virtual machines, storage, databases, identities and other cloud assets. | Find internet-exposed virtual machines. |
| Security condition | Vulnerabilities, recommendations, misconfiguration and posture findings. | Find resources with critical vulnerabilities. |
| Network relationship | Public exposure, connectivity and reachable resources. | Find vulnerable assets reachable from the internet. |
| Identity relationship | Users, service principals, managed identities and permissions. | Find identities with high privilege over exposed resources. |
| Data relationship | Connections between compute, storage, databases and secrets. | Find exposed applications connected to sensitive data stores. |
| Cloud scope | Azure subscriptions, AWS accounts and GCP projects. | Compare exposure across multiple cloud environments. |
| Tags and ownership | Business unit, application, environment and owner metadata. | Find production assets without an assigned owner. |
| Recommendation state | Open, exempted, unhealthy or unresolved findings. | Find critical recommendations on internet-facing resources. |
Useful investigation questions
- Which internet-facing resources have critical vulnerabilities?
- Which storage accounts are publicly accessible and contain sensitive data?
- Which identities have excessive permissions over production resources?
- Which virtual machines are exposed and missing endpoint protection?
- Which applications can reach critical databases?
- Which resources have high-risk recommendations and no owner?
- Which cloud assets are connected to privileged identities?
- Which exempted recommendations still affect externally exposed resources?
- Which resources combine public exposure, weak identity controls and known vulnerabilities?
- Which multicloud assets create the greatest potential blast radius?
Agent Foskett investigation: “The database was private… but the application was not.”
Key takeaways
- Cloud Security Explorer investigates relationships rather than isolated findings.
- Queries can combine resource type, exposure, vulnerabilities, permissions and ownership.
- Public exposure becomes more serious when combined with exploitable weaknesses.
- Identity relationships can reveal hidden access paths.
- Private resources may still be reachable through exposed applications or overprivileged identities.
- Cloud Security Explorer supports Azure, AWS and GCP visibility where connected.
- Saved investigation patterns can support repeatable posture reviews.
- Business context remains essential when interpreting query results.
- Findings should be validated before remediation decisions are made.
- The most important risk is often the combination of several smaller weaknesses.
Learning objectives
After completing this lesson, you should be able to build Cloud Security Explorer queries, analyse relationships and identify risky resource combinations.
What is Cloud Security Explorer?
Cloud Security Explorer is a query-based investigation capability for analysing cloud assets and security relationships.
Why relationships matter
Attackers move through permissions, connectivity and dependencies rather than treating resources as isolated objects.
Query builder
The query builder allows investigators to combine resource conditions and relationships without writing code.
Resource conditions
Resource conditions filter assets by type, state, configuration, recommendation or exposure.
Relationship conditions
Relationship conditions connect resources to identities, networks, data stores and other assets.
Azure resources
Explorer can investigate supported Azure resources across subscriptions and management scopes.
AWS resources
Connected AWS accounts can contribute supported asset and posture relationships.
GCP resources
Connected GCP projects can contribute supported asset and posture relationships.
Public exposure
Publicly reachable resources should be examined alongside vulnerabilities and permissions.
Critical vulnerabilities
Critical vulnerabilities become more urgent when the affected resource is reachable or highly privileged.
Identity exposure
Identity relationships can reveal users, service principals and managed identities with dangerous access.
Overprivileged identities
Excessive permissions can turn a small compromise into a much larger incident.
Managed identities
Managed identities should be reviewed for unnecessary access to sensitive resources.
Service principals
Service principals can create hidden access paths when credentials or permissions are poorly controlled.
Storage exposure
Public or broadly accessible storage should be investigated for sensitive content and linked applications.
Database relationships
Databases may be indirectly exposed through connected applications and identities.
Network reachability
Reachability helps determine whether one resource can communicate with another.
Attack paths
Attack paths combine relationships into a route an attacker could potentially exploit.
Recommendations
Explorer queries can use security recommendations as investigation conditions.
Exemptions
Exempted findings should still be reviewed when they participate in a risky exposure chain.
Ownership
Ownership metadata helps route findings to the correct team.
Tags
Tags can help separate production, development, business-unit and application resources.
Environment filters
Environment filters can narrow investigations to subscriptions, accounts, projects or regions.
Risk combinations
Several moderate weaknesses can combine into one high-impact exposure.
Blast radius
Blast radius describes how far an attacker could move after compromising a resource or identity.
Lateral movement
Relationships can reveal possible movement between applications, identities and data.
Sensitive resources
Critical databases, secrets, key stores and privileged systems deserve additional context.
Saved queries
Repeatable query patterns help security teams perform consistent posture reviews.
Investigation workflow
Start with a question, build the query, inspect results, validate context and prioritise action.
Result validation
Confirm that each relationship and condition accurately reflects the environment.
False assumptions
Do not assume that private, internal or low-severity resources are automatically low risk.
Prioritisation
Prioritise combinations that include public exposure, privilege, sensitive data and exploitable weaknesses.
Remediation planning
Break the risky path by removing one or more critical relationships or conditions.
Patch management
Patching can remove the exploitable weakness from an otherwise risky chain.
Permission reduction
Least privilege can break identity-based exposure paths.
Network controls
Segmentation and private access can reduce reachability.
Resource isolation
Isolation can contain a high-risk resource while permanent remediation is completed.
Multicloud visibility
Use consistent investigation questions across connected Azure, AWS and GCP environments.
Reporting
Summarise the affected assets, relationships, business impact and remediation priorities.
Collaboration
Security, platform, identity and application teams may all be needed to resolve complex exposure paths.
Continuous review
Rerun important queries because cloud relationships change frequently.
New resources
New deployments can introduce fresh exposure and permission relationships.
Configuration drift
Configuration changes can recreate previously resolved risky combinations.
Query scope
Begin with a focused scope, then expand only when broader investigation is required.
Evidence
Retain query results, affected resource details and remediation decisions.
What Agent Foskett checked
Agent Foskett checked public exposure, vulnerabilities, managed identity permissions and database relationships.
Best practices
- Begin with a clear investigation question.
- Combine exposure, vulnerability, identity and data conditions.
- Use the smallest practical scope.
- Validate query results against the actual environment.
- Prioritise risky combinations rather than isolated findings.
- Review identity permissions and network reachability together.
- Save useful investigation patterns.
- Rerun queries after major cloud changes.
- Document business context and ownership.
- Break the exposure path through targeted remediation.
Related Agent Foskett resources
Continue learning
What is Microsoft Defender for Cloud Security Explorer?
Microsoft Defender for Cloud Security Explorer is a query-based investigation capability that helps security teams analyse cloud assets, identities, permissions, connectivity and security relationships.
Cloud Security Explorer Lesson
This Agent Foskett lesson explains asset queries, relationship conditions, exposure analysis, identity permissions, network reachability, attack paths, multicloud visibility and risky resource combinations.
