LAB 19: Microsoft Sentinel for Identity Security

# Create Log Analytics Workspace via CLI az monitor log-analytics workspace create \ --resource-group rg-security \ --workspace-name law-sentinel-prod \ --location eastus \ --retention-time 90 \ --sku PerGB2018 # Or via Portal: # Azure Portal Log Analytics workspaces Create # - Subscription: Select subscription # - Resource group: rg-security # - Name: law-sentinel-prod # - Region: East US (or your region) # Review + Create # Retention Settings: # - Default: 30 days (free) # - Recommended: 90-365 days for security # - Compliance may require longer

Navigate to: Azure Portal Microsoft Sentinel Add Sentinel to Workspace: Click "Add" Select your Log Analytics workspace (law-sentinel-prod) Click "Add" Initial Setup: - Sentinel solution deployed on workspace - Content hub available - Data connectors gallery enabled Pricing: - Pay-per-GB ingested - Commitment tiers available (100GB+) - Free tier: 10GB/day first 31 days First 90 Days: - Microsoft 365 E5 includes 5GB/user/month - Entra ID P2 logs free to ingest - Plan capacity based on log volume

Navigate to: Sentinel Data connectors Search "Microsoft Entra ID" Open connector page Configure Connection: SIGN-IN LOGS: ☑ Azure AD Sign-in logs - Contains: All user sign-ins - Table: SigninLogs ☑ Azure AD Non-Interactive Sign-in logs - Contains: Service principal, background auth - Table: AADNonInteractiveUserSignInLogs ☑ Azure AD Service Principal Sign-in logs - Contains: App/service authentications - Table: AADServicePrincipalSignInLogs ☑ Azure AD Managed Identity Sign-in logs - Contains: Managed identity authentications - Table: AADManagedIdentitySignInLogs AUDIT LOGS: ☑ Azure AD Audit logs - Contains: Directory changes, PIM, CA - Table: AuditLogs ☑ Azure AD Provisioning logs - Contains: User provisioning events - Table: AADProvisioningLogs Apply Changes Verify: Run query after 5-10 minutes SigninLogs | take 10

Microsoft 365 Defender Connector: Data connectors Microsoft 365 Defender Enable: ☑ Microsoft Defender for Identity - On-premises AD monitoring - Lateral movement detection - Table: IdentityLogonEvents, IdentityDirectoryEvents ☑ Microsoft Defender for Cloud Apps - Cloud app usage - Shadow IT - Table: CloudAppEvents ☑ Microsoft Defender for Office 365 - Email threats - Phishing attempts - Table: EmailEvents --- Identity Protection (Entra ID P2): Data connectors Azure AD Identity Protection Enable: ☑ Security Alerts - Risk detections - Table: SecurityAlert (filtered) ☑ Risky Users - Users flagged as risky - Table: AADRiskyUsers ☑ User Risk Events - Individual risk events - Table: AADUserRiskEvents

// KQL Query Structure TableName | where TimeGenerated > ago(24h) // Filter by time | where Column == "Value" // Filter rows | project Column1, Column2, Column3 // Select columns | summarize count() by Column1 // Aggregate | order by count_ desc // Sort | take 10 // Limit results // Common Operators | where // Filter rows | project // Select/rename columns | extend // Add calculated columns | summarize // Aggregate (count, sum, avg, etc.) | join // Combine tables | union // Append tables | order by // Sort results | take / limit // Limit rows | render // Visualize (timechart, piechart, etc.) // Time Functions ago(1h) // 1 hour ago ago(7d) // 7 days ago ago(30d) // 30 days ago now() // Current time startofday() // Start of today between(datetime..datetime) // String Functions contains // Case-insensitive contains has // Word boundary match (faster) startswith // Starts with matches regex // Regex match tolower() // Lowercase tostring() // Convert to string

// Failed Sign-ins by User (Last 24 Hours) SigninLogs | where TimeGenerated > ago(24h) | where ResultType != 0 // Non-success | summarize FailedCount = count(), DistinctIPs = dcount(IPAddress), Apps = make_set(AppDisplayName) by UserPrincipalName | where FailedCount > 10 | order by FailedCount desc // Sign-ins from New Countries (First Time Ever) let KnownCountries = SigninLogs | where TimeGenerated between (ago(90d) .. ago(1d)) | distinct UserPrincipalName, Location; SigninLogs | where TimeGenerated > ago(1d) | join kind=leftanti KnownCountries on UserPrincipalName, Location | project TimeGenerated, UserPrincipalName, Location, IPAddress, AppDisplayName, ResultType // Successful Sign-ins After Multiple Failures SigninLogs | where TimeGenerated > ago(1h) | summarize FailCount = countif(ResultType != 0), SuccessCount = countif(ResultType == 0), LastResult = arg_max(TimeGenerated, ResultType) by UserPrincipalName, IPAddress | where FailCount > 5 and SuccessCount > 0 | project UserPrincipalName, IPAddress, FailCount, SuccessCount // MFA Bypass Attempts (Legacy Auth) SigninLogs | where TimeGenerated > ago(24h) | where ClientAppUsed in ("Exchange ActiveSync", "IMAP4", "POP3", "SMTP", "Other clients") | summarize count() by UserPrincipalName, ClientAppUsed | order by count_ desc

// PIM Role Activations AuditLogs | where TimeGenerated > ago(7d) | where OperationName has "PIM" and OperationName has "activated" | extend User = tostring(InitiatedBy.user.userPrincipalName) | extend Role = tostring(TargetResources[0].displayName) | project TimeGenerated, User, Role, OperationName | order by TimeGenerated desc // Global Admin Role Assignments AuditLogs | where TimeGenerated > ago(30d) | where OperationName == "Add member to role" | extend Role = tostring(TargetResources[0].modifiedProperties[1].newValue) | where Role has "Global Administrator" or Role has "62e90394-69f5-4237-9190-012177145e10" | extend AssignedUser = tostring(TargetResources[0].userPrincipalName) | extend AssignedBy = tostring(InitiatedBy.user.userPrincipalName) | project TimeGenerated, AssignedUser, AssignedBy, Role // Privileged Role Members (Current) IdentityInfo | where TimeGenerated > ago(1d) | where AssignedRoles has "Global Administrator" or AssignedRoles has "Privileged Role Administrator" | project AccountUPN, AssignedRoles, Department, JobTitle // Service Principal Permission Changes AuditLogs | where TimeGenerated > ago(7d) | where OperationName in ( "Add app role assignment to service principal", "Add delegated permission grant", "Consent to application") | extend AppName = tostring(TargetResources[0].displayName) | extend Permission = tostring(TargetResources[0].modifiedProperties) | extend Actor = tostring(InitiatedBy.user.userPrincipalName) | project TimeGenerated, Actor, AppName, OperationName, Permission

// High-Risk Sign-ins SigninLogs | where TimeGenerated > ago(24h) | where RiskLevelDuringSignIn in ("high", "medium") | project TimeGenerated, UserPrincipalName, RiskLevelDuringSignIn, RiskEventTypes, IPAddress, Location, AppDisplayName | order by TimeGenerated desc // Users with Multiple Risk Detections AADUserRiskEvents | where TimeGenerated > ago(30d) | summarize RiskCount = count(), RiskTypes = make_set(RiskEventType), LastDetection = max(TimeGenerated) by UserPrincipalName | where RiskCount > 3 | order by RiskCount desc // Impossible Travel Detection SigninLogs | where TimeGenerated > ago(24h) | where ResultType == 0 | summarize Locations = make_set(Location), LocationCount = dcount(Location), SigninTimes = make_list(TimeGenerated) by UserPrincipalName, bin(TimeGenerated, 1h) | where LocationCount > 1 | project UserPrincipalName, Locations, SigninTimes // Anonymous IP Sign-ins SigninLogs | where TimeGenerated > ago(7d) | where RiskEventTypes has "anonymizedIPAddress" or IPAddress in ( // Known Tor exit nodes (example) "185.220.101.0/24" ) | project TimeGenerated, UserPrincipalName, IPAddress, Location

// User Enumeration Attempts SigninLogs | where TimeGenerated > ago(1h) | where ResultType == 50034 // User not found | summarize AttemptedUsers = dcount(UserPrincipalName), SourceIPs = make_set(IPAddress) by bin(TimeGenerated, 5m) | where AttemptedUsers > 10 // Password Spray Detection SigninLogs | where TimeGenerated > ago(1h) | where ResultType in (50126, 50053) // Invalid password, locked | summarize TargetedUsers = dcount(UserPrincipalName), Attempts = count() by IPAddress, bin(TimeGenerated, 5m) | where TargetedUsers > 10 | project TimeGenerated, IPAddress, TargetedUsers, Attempts // Group Membership Reconnaissance AuditLogs | where TimeGenerated > ago(24h) | where OperationName in ( "Get group members", "Get group", "List groups") | extend Actor = tostring(InitiatedBy.user.userPrincipalName) | summarize QueryCount = count() by Actor, bin(TimeGenerated, 1h) | where QueryCount > 50 // Mass Download of User List AuditLogs | where TimeGenerated > ago(24h) | where OperationName == "Download users" | extend Actor = tostring(InitiatedBy.user.userPrincipalName) | project TimeGenerated, Actor, OperationName

// Conditional Access Failures SigninLogs | where TimeGenerated > ago(24h) | mv-expand ConditionalAccessPolicies | extend PolicyName = tostring(ConditionalAccessPolicies.displayName) | extend PolicyResult = tostring(ConditionalAccessPolicies.result) | where PolicyResult == "failure" | summarize count() by PolicyName, UserPrincipalName | order by count_ desc // Sign-ins Bypassing All Policies SigninLogs | where TimeGenerated > ago(24h) | where ResultType == 0 // Successful | mv-expand ConditionalAccessPolicies | extend PolicyResult = tostring(ConditionalAccessPolicies.result) | summarize AppliedPolicies = countif(PolicyResult == "success"), NotApplied = countif(PolicyResult == "notApplied") by UserPrincipalName, IPAddress | where AppliedPolicies == 0 and NotApplied > 0 | project UserPrincipalName, IPAddress, NotApplied // Conditional Access Policy Changes AuditLogs | where TimeGenerated > ago(7d) | where OperationName has "conditional access" | extend PolicyName = tostring(TargetResources[0].displayName) | extend Actor = tostring(InitiatedBy.user.userPrincipalName) | project TimeGenerated, Actor, OperationName, PolicyName

Analytics Rule Types: 1. SCHEDULED RULES - Run KQL query on schedule (5min - 14 days) - Most common type - Full customization 2. NRT (Near Real-Time) RULES - Run every minute - 30-second latency - For critical detections 3. FUSION RULES - ML-based correlation - Built-in, Microsoft-managed - Multi-stage attack detection 4. MICROSOFT SECURITY RULES - Import alerts from other products - Defender for Identity, Cloud Apps, etc. 5. ANOMALY RULES - ML-based anomaly detection - UEBA-powered - Customizable thresholds Navigate to: Sentinel Analytics Rule templates (pre-built) Active rules (currently enabled) Create Scheduled query rule

// Navigate to: Analytics Create Scheduled query rule // GENERAL TAB Name: Identity - Brute Force Attack Detected Description: Detects multiple failed sign-ins followed by success Severity: High MITRE ATT&CK: Credential Access - Brute Force (T1110) Status: Enabled // SET RULE LOGIC TAB // Query: let threshold = 10; let timeframe = 1h; SigninLogs | where TimeGenerated > ago(timeframe) | summarize FailedCount = countif(ResultType != 0), SuccessCount = countif(ResultType == 0), IPAddresses = make_set(IPAddress), Apps = make_set(AppDisplayName), FailureCodes = make_set(ResultType) by UserPrincipalName | where FailedCount > threshold and SuccessCount > 0 | extend AccountName = tostring(split(UserPrincipalName, "@")[0]) | extend AccountDomain = tostring(split(UserPrincipalName, "@")[1]) | project UserPrincipalName, AccountName, AccountDomain, FailedCount, SuccessCount, IPAddresses, Apps // Entity Mapping: // - Account UserPrincipalName (Name), AccountDomain (UPNSuffix) // - IP IPAddresses // Query Scheduling: // - Run every: 1 hour // - Lookup data from: 1 hour // Alert Threshold: // - Generate alert when: Number of results > 0 // Event Grouping: // - Group events into single alert

// Rule: Global Admin PIM Activation Outside Business Hours // GENERAL TAB Name: Identity - Off-Hours Global Admin Activation Description: Global Admin role activated outside business hours Severity: High MITRE ATT&CK: Privilege Escalation - Valid Accounts (T1078) // QUERY: AuditLogs | where TimeGenerated > ago(1h) | where OperationName has "Add member to role in PIM completed" | extend Role = tostring(TargetResources[0].displayName) | where Role has "Global Administrator" | extend User = tostring(InitiatedBy.user.userPrincipalName) | extend Hour = hourofday(TimeGenerated) | extend DayOfWeek = dayofweek(TimeGenerated) // Outside 7 AM - 7 PM or weekends | where Hour < 7 or Hour >19 or DayOfWeek in (0d, 6d) | project TimeGenerated, User, Role, Hour, DayOfWeek | extend AccountName = tostring(split(User, "@")[0]) | extend AccountDomain = tostring(split(User, "@")[1]) // Entity Mapping: // - Account User // Scheduling: // - Run every: 1 hour // - Lookup: 1 hour // Incident Settings: // - Create incidents: Yes // - Group related alerts: Yes

// Navigate to: Analytics Create NRT query rule // GENERAL TAB Name: Identity - Sign-in from New Country (NRT) Description: User signed in from country not seen in past 14 days Severity: Medium MITRE ATT&CK: Initial Access - Valid Accounts (T1078) // QUERY: let KnownLocations = SigninLogs | where TimeGenerated between (ago(14d) .. ago(1d)) | where ResultType == 0 | distinct UserPrincipalName, Location; SigninLogs | where TimeGenerated > ago(10m) // NRT queries recent data | where ResultType == 0 | join kind=leftanti KnownLocations on UserPrincipalName, Location | project TimeGenerated, UserPrincipalName, Location, IPAddress, AppDisplayName, DeviceDetail, UserAgent | extend AccountName = tostring(split(UserPrincipalName, "@")[0]) | extend AccountDomain = tostring(split(UserPrincipalName, "@")[1]) // Entity Mapping: // - Account UserPrincipalName // - IP IPAddress // - Host (if available) // NRT runs every minute automatically

Navigate to: Sentinel Settings Settings tab User and Entity Behavior Analytics ENABLE UEBA: ☑ Enable UEBA DATA SOURCES: ☑ Azure Active Directory (Entra ID) ☑ Azure Activity ☑ Security Events (Windows) ☑ Office Activity ENTITY SYNC: ☑ Sync Entra ID entities - Syncs users, groups, devices - Builds identity profiles Apply UEBA Features Enabled: - Entity pages (user profiles) - Behavioral baselines - Anomaly scoring - Investigation priority - Timeline view Wait 24-48 hours for baseline to build

Navigate to: Sentinel Entity behavior Search for User: Search "john@contoso.com" Click user entity USER ENTITY PAGE SECTIONS: 1. OVERVIEW - Display name, UPN, job title - Department, manager - Risk score (if available) - Investigation priority 2. TIMELINE - All activities chronologically - Sign-ins, audit events - Anomalies highlighted 3. INSIGHTS - First time activities - Anomalous behaviors - Peer group comparison 4. ALERTS - Alerts involving this user - Related incidents 5. RELATED ENTITIES - Groups membership - Devices used - IPs connected from Investigation Priority Score: - 0-3: Low priority - 4-6: Medium priority - 7-10: High priority (investigate immediately)

// Query UEBA Anomalies BehaviorAnalytics | where TimeGenerated > ago(24h) | where ActivityInsights has "True" // Anomaly detected | project TimeGenerated, UserPrincipalName, ActivityType, ActionType, ActivityInsights, InvestigationPriority, SourceIPAddress, SourceDevice | order by InvestigationPriority desc // High Priority Anomalies Only BehaviorAnalytics | where TimeGenerated > ago(7d) | where InvestigationPriority > 5 | summarize AnomalyCount = count(), Activities = make_set(ActivityType) by UserPrincipalName | order by AnomalyCount desc // First Time User Activity BehaviorAnalytics | where TimeGenerated > ago(24h) | where ActivityInsights has "FirstTimeUserUsedApp" or ActivityInsights has "FirstTimeUserConnectedFromCountry" | project TimeGenerated, UserPrincipalName, ActivityType, ActivityInsights // Anomalous Resource Access BehaviorAnalytics | where TimeGenerated > ago(7d) | where ActionType == "ResourceAccess" | where ActivityInsights has "AnomalousResource" | project TimeGenerated, UserPrincipalName, DestinationResource, ActivityInsights

// Analytics Rule: High Priority UEBA Anomaly // Create Scheduled Rule Name: UEBA - High Investigation Priority User Severity: Medium MITRE: Discovery - Account Discovery (T1087) // Query: BehaviorAnalytics | where TimeGenerated > ago(1h) | where InvestigationPriority >= 7 | summarize AnomalyTypes = make_set(ActivityInsights), Activities = make_set(ActivityType), Devices = make_set(SourceDevice), IPs = make_set(SourceIPAddress) by UserPrincipalName | extend AccountName = tostring(split(UserPrincipalName, "@")[0]) | extend AccountDomain = tostring(split(UserPrincipalName, "@")[1]) --- // Analytics Rule: User Accessing Unusual Resource Name: UEBA - Anomalous Resource Access Severity: Low // Query: BehaviorAnalytics | where TimeGenerated > ago(1h) | where ActivityInsights has "FirstTimeUserAccessedResource" or ActivityInsights has "ResourceUncommonlyAccessedByUser" | project TimeGenerated, UserPrincipalName, DestinationResource, ActivityInsights, InvestigationPriority | where InvestigationPriority > 3

Navigate to: Sentinel Incidents INCIDENT QUEUE: - List of all incidents - Sortable by severity, status, time - Filterable by owner, product, etc. INCIDENT PROPERTIES: - Title: From analytics rule name - Severity: High/Medium/Low/Informational - Status: New Active Closed - Owner: Assigned analyst - Product: Source (Sentinel, Defender, etc.) TRIAGE WORKFLOW: 1. Review new incidents 2. Assess severity and validity 3. Assign owner 4. Set status to "Active" 5. Investigate 6. Respond/Remediate 7. Close with classification CLASSIFICATION OPTIONS: - True Positive - Suspicious activity - Benign Positive - Expected behavior - False Positive - Incorrect detection - Undetermined BULK OPERATIONS: - Select multiple incidents - Change status - Assign owner - Add tags

Open Incident Click "Full details" INVESTIGATION TAB: Evidence: - Alerts that created incident - Events from query results - Bookmarks added Entities: - Users involved - IP addresses - Hosts/devices - Applications Timeline: - Chronological view of events - Related activities - Entity actions Comments: - Add investigation notes - Collaboration with team - Audit trail --- INVESTIGATION GRAPH: Click "Investigate" Visual representation: - Central entity (user/IP) - Related entities connected - Expand to see relationships - Click entity for details Entity Expansion: - User Sign-ins, groups, devices - IP Other users from same IP - Host Users who logged in Useful for: - Understanding attack scope - Finding lateral movement - Identifying all affected entities

// Investigate Specific User (Last 24 Hours) let TargetUser = "compromised.user@contoso.com"; union SigninLogs, AuditLogs | where TimeGenerated > ago(24h) | where UserPrincipalName == TargetUser or InitiatedBy.user.userPrincipalName == TargetUser | project TimeGenerated, Type, OperationName, ResultType, IPAddress, Location | order by TimeGenerated desc // All Activity from Suspicious IP let SuspiciousIP = "203.0.113.50"; SigninLogs | where TimeGenerated > ago(7d) | where IPAddress == SuspiciousIP | project TimeGenerated, UserPrincipalName, AppDisplayName, ResultType, Location // User's Device History SigninLogs | where TimeGenerated > ago(30d) | where UserPrincipalName == "user@contoso.com" | extend DeviceName = tostring(DeviceDetail.displayName) | extend DeviceOS = tostring(DeviceDetail.operatingSystem) | distinct DeviceName, DeviceOS, DeviceDetail // Actions Taken by Compromised Account AuditLogs | where TimeGenerated > ago(7d) | where InitiatedBy.user.userPrincipalName == "compromised@contoso.com" | project TimeGenerated, OperationName, TargetResources, Result | order by TimeGenerated desc

Identity Incident Response Checklist: IMMEDIATE ACTIONS (Containment): ☐ Disable compromised account ☐ Revoke all sessions (Entra ID) ☐ Reset password ☐ Revoke refresh tokens ☐ Block suspicious IP (Conditional Access) INVESTIGATION: ☐ Determine scope (how many accounts?) ☐ Identify initial access vector ☐ Check for persistence mechanisms ☐ Review privileged access changes ☐ Check for data exfiltration ☐ Identify affected applications PERSISTENCE CHECK: ☐ New MFA devices registered? ☐ New OAuth applications consented? ☐ New credentials on service principals? ☐ New mail forwarding rules? ☐ New admin role assignments? ERADICATION: ☐ Remove malicious OAuth apps ☐ Remove unauthorized MFA devices ☐ Remove mail forwarding rules ☐ Remove added credentials ☐ Revert privilege escalations RECOVERY: ☐ Re-enable account with new password ☐ Re-register MFA ☐ Educate user on phishing ☐ Monitor for re-compromise POST-INCIDENT: ☐ Document timeline ☐ Update detection rules ☐ Lessons learned ☐ Close incident with classification

Sentinel Playbooks = Azure Logic Apps Components: 1. TRIGGER - When incident is created - When alert is triggered - Manual trigger 2. CONNECTOR ACTIONS - Entra ID (disable user, revoke sessions) - Microsoft Teams (post message) - Office 365 (send email) - Microsoft Graph (API calls) - ServiceNow (create ticket) 3. LOGIC - Conditions (if/then) - Loops (for each entity) - Variables Prerequisites: - Logic App Contributor role - Appropriate permissions on target systems - Managed Identity or service principal Navigate to: Sentinel Automation Playbook templates (pre-built) Active playbooks (configured) Create Playbook (Logic App)

Create Playbook: Disable-CompromisedUser 1. CREATE LOGIC APP: Navigate to: Sentinel Automation Create Playbook - Name: Disable-CompromisedUser - Resource group: rg-security - Enable managed identity: Yes 2. TRIGGER: - Microsoft Sentinel incident (When incident is created) - Or: Microsoft Sentinel alert 3. GET INCIDENT ENTITIES: - Action: Entities - Get Accounts - Incident ARM ID: From trigger 4. FOR EACH ACCOUNT: - Loop through accounts 5. DISABLE USER (Azure AD Connector): - Action: Update user - User ID: Account object ID - Account Enabled: false 6. REVOKE SESSIONS (HTTP Action): POST https://graph.microsoft.com/v1.0/users/{id}/revokeSignInSessions - Authentication: Managed Identity - Audience: https://graph.microsoft.com 7. POST TO TEAMS: - Action: Post message (Teams) - Channel: Security-Alerts - Message: "User [UPN] disabled due to incident [ID]" 8. ADD COMMENT TO INCIDENT: - Action: Add comment to incident - Comment: "Automated response: User disabled, sessions revoked" PERMISSIONS NEEDED: - User.ReadWrite.All (disable user) - User.RevokeSessions.All (revoke sessions)

// Playbook: Block-MaliciousIP // This playbook adds IP to a Named Location (blocklist) // which is used in Conditional Access to block sign-ins // PRE-REQUISITES: // 1. Create Named Location in Entra ID: // - Name: "Blocked-IPs-Sentinel" // - Type: IP ranges // - Initially empty // 2. Create CA Policy: // - Condition: Location = "Blocked-IPs-Sentinel" // - Access: Block // LOGIC APP FLOW: // 1. Trigger: Microsoft Sentinel incident // 2. Get IPs: // Action: Entities - Get IPs // Input: Incident ARM ID // 3. For Each IP: // 4. Get Current Named Location (HTTP): GET https://graph.microsoft.com/v1.0/identity/conditionalAccess/namedLocations/{id} // 5. Add IP to list: // Parse current IPs, append new IP // 6. Update Named Location (HTTP): PATCH https://graph.microsoft.com/v1.0/identity/conditionalAccess/namedLocations/{id} { "@odata.type": "#microsoft.graph.ipNamedLocation", "ipRanges": [ {"@odata.type": "#microsoft.graph.iPv4CidrRange", "cidrAddress": "x.x.x.x/32"} ] } // 7. Add incident comment: // "IP x.x.x.x added to block list"

Playbook: Enrich-UserDetails Purpose: Add user context to incident for faster triage FLOW: 1. Trigger: When incident created 2. Get account entities from incident 3. For each account: 4. Get User Details (Graph API): GET /users/{id}?$select=displayName,jobTitle, department,manager,accountEnabled 5. Get User Risk (Graph API): GET /identityProtection/riskyUsers/{id} 6. Get Recent Sign-ins (Graph API): GET /auditLogs/signIns?$filter=userId eq '{id}' &$top=5&$orderby=createdDateTime desc 7. Get Group Memberships (Graph API): GET /users/{id}/memberOf 8. Build enrichment summary: - Name: John Doe - Title: Finance Manager - Department: Finance - Manager: Jane Smith - Risk Level: High - Recent Sign-ins: 5 from US, 1 from Russia (!) - Privileged Groups: None 9. Add comment to incident with enrichment 10. Update incident description with context Benefits: - Analyst has context immediately - Faster triage decisions - Consistent enrichment process

Attach Playbook to Rule: Option 1: Via Analytics Rule Analytics Select rule Edit Automated response tab Add playbook Select playbook (must have incident trigger) Option 2: Via Automation Rules Automation Create Automation rule Automation Rule Configuration: - Name: Auto-respond to brute force - Trigger: When incident is created - Conditions: - Analytic rule name Contains "Brute Force" - Severity Equals High - Actions: - Run playbook: Disable-CompromisedUser - Change status: Active - Assign owner: SOC Team Automation Rule Benefits: - Apply playbooks to multiple rules - Conditional execution - Change incident properties - Route to specific teams

Navigate to: Sentinel Workbooks Templates IDENTITY WORKBOOKS: 1. Azure AD Sign-in Logs - Sign-in trends - Failure analysis - MFA statistics - Conditional Access impact 2. Azure AD Audit Logs - Directory changes - Role assignments - Application changes 3. Insecure Protocols - Legacy authentication usage - Users at risk - Apps using legacy auth 4. Identity and Access - Privileged users overview - Risky sign-ins - Risk detections 5. User and Entity Behavior Analytics - Anomaly trends - Top risky users - Investigation priorities INSTALL WORKBOOK: Select template Save (creates copy in your workspace) View saved workbooks

// Navigate to: Workbooks Add workbook Edit // Add Parameter: Time Range // Type: Time range picker // Default: Last 24 hours // ========== TILE 1: Sign-in Summary ========== // Add Query Paste: SigninLogs | where TimeGenerated {TimeRange} | summarize TotalSignins = count(), SuccessfulSignins = countif(ResultType == 0), FailedSignins = countif(ResultType != 0), UniqueUsers = dcount(UserPrincipalName) | project TotalSignins, SuccessfulSignins, FailedSignins, UniqueUsers // Visualization: Tiles // ========== TILE 2: Failed Sign-ins Over Time ========== SigninLogs | where TimeGenerated {TimeRange} | where ResultType != 0 | summarize FailedCount = count() by bin(TimeGenerated, 1h) | render timechart // ========== TILE 3: Top Failed Users ========== SigninLogs | where TimeGenerated {TimeRange} | where ResultType != 0 | summarize FailCount = count() by UserPrincipalName | top 10 by FailCount | render barchart // ========== TILE 4: Risky Sign-ins ========== SigninLogs | where TimeGenerated {TimeRange} | where RiskLevelDuringSignIn in ("high", "medium") | summarize count() by RiskLevelDuringSignIn | render piechart // ========== TILE 5: Sign-ins by Location ========== SigninLogs | where TimeGenerated {TimeRange} | where ResultType == 0 | summarize count() by Location | top 10 by count_ | render piechart

// PIM Monitoring Dashboard // ========== PIM Activations Over Time ========== AuditLogs | where TimeGenerated {TimeRange} | where OperationName has "PIM" and OperationName has "activated" | summarize Activations = count() by bin(TimeGenerated, 4h) | render timechart // ========== Activations by Role ========== AuditLogs | where TimeGenerated {TimeRange} | where OperationName has "PIM" and OperationName has "activated" | extend Role = tostring(TargetResources[0].displayName) | summarize count() by Role | render piechart // ========== Top Users Activating Roles ========== AuditLogs | where TimeGenerated {TimeRange} | where OperationName has "PIM" and OperationName has "activated" | extend User = tostring(InitiatedBy.user.userPrincipalName) | summarize ActivationCount = count() by User | top 10 by ActivationCount | render barchart // ========== Off-Hours Activations ========== AuditLogs | where TimeGenerated {TimeRange} | where OperationName has "PIM" and OperationName has "activated" | extend Hour = hourofday(TimeGenerated) | extend OffHours = iff(Hour < 7 or Hour > 19, "Off-Hours", "Business Hours") | summarize count() by OffHours | render piechart // ========== Recent Activations Table ========== AuditLogs | where TimeGenerated {TimeRange} | where OperationName has "PIM" and OperationName has "activated" | extend User = tostring(InitiatedBy.user.userPrincipalName) | extend Role = tostring(TargetResources[0].displayName) | project TimeGenerated, User, Role, OperationName | order by TimeGenerated desc | take 50

Sharing Workbooks: SAVE OPTIONS: Save Save as - Save to: My Workbooks (personal) or Shared (team) - Name: Identity Security Dashboard - Subscription/Resource Group: Select SHARE WITH TEAM: - Shared workbooks visible to all with Reader access - Can pin to Azure Dashboard EXPORT OPTIONS: More Export - Download as .workbook file - Import to other workspaces PIN TO AZURE DASHBOARD: - Individual tiles can be pinned - Creates quick-access dashboard SCHEDULED REPORTS: Currently requires: 1. Logic App with recurrence trigger 2. Run KQL query via Log Analytics API 3. Format results 4. Email via Office 365 connector Or use: - Power BI with Log Analytics connector - Scheduled refresh and email subscriptions

Navigate to: Sentinel Hunting HUNTING FEATURES: - Pre-built queries (templates) - Custom queries - Bookmarks (save findings) - Livestream (real-time) HUNTING WORKFLOW: 1. Form hypothesis 2. Write/select query 3. Run query 4. Analyze results 5. Bookmark interesting findings 6. Create incident if threat confirmed HYPOTHESIS EXAMPLES: - "Attackers may have added persistence via OAuth apps" - "Compromised accounts may be accessing unusual resources" - "Lateral movement via PIM abuse" BUILT-IN IDENTITY HUNTING QUERIES: Hunting Queries Filter by "Identity" Categories: - Credential Access - Persistence - Privilege Escalation - Initial Access - Discovery

// Hunt for malicious OAuth app persistence // New OAuth Apps with Mail.Read Permission AuditLogs | where TimeGenerated > ago(30d) | where OperationName == "Consent to application" | extend AppName = tostring(TargetResources[0].displayName) | extend Permissions = tostring(TargetResources[0].modifiedProperties) | where Permissions has "Mail.Read" or Permissions has "Mail.ReadWrite" | extend ConsentedBy = tostring(InitiatedBy.user.userPrincipalName) | project TimeGenerated, ConsentedBy, AppName, Permissions // Apps with High-Risk Permissions AuditLogs | where TimeGenerated > ago(30d) | where OperationName in ( "Add delegated permission grant", "Add app role assignment to service principal") | extend Permissions = tostring(TargetResources[0].modifiedProperties) | where Permissions has_any ( "Directory.ReadWrite.All", "Mail.ReadWrite", "Files.ReadWrite.All", "RoleManagement.ReadWrite.Directory") | extend AppName = tostring(TargetResources[0].displayName) | extend Actor = tostring(InitiatedBy.user.userPrincipalName) | project TimeGenerated, Actor, AppName, Permissions // Service Principal Credentials Added AuditLogs | where TimeGenerated > ago(30d) | where OperationName in ( "Add service principal credentials", "Update application – Certificates and secrets management") | extend AppName = tostring(TargetResources[0].displayName) | extend Actor = tostring(InitiatedBy.user.userPrincipalName) | project TimeGenerated, Actor, AppName, OperationName

// Hunt for privilege escalation attempts // Users Added to Privileged Groups AuditLogs | where TimeGenerated > ago(7d) | where OperationName == "Add member to group" | extend GroupName = tostring(TargetResources[0].displayName) | where GroupName has_any ( "Global Administrator", "Privileged", "Admin", "Domain Admins", "Enterprise Admins") | extend AddedUser = tostring(TargetResources[1].userPrincipalName) | extend AddedBy = tostring(InitiatedBy.user.userPrincipalName) | project TimeGenerated, AddedBy, AddedUser, GroupName // Direct Role Assignments (Bypassing PIM) AuditLogs | where TimeGenerated > ago(7d) | where OperationName == "Add member to role" | where OperationName !has "PIM" // Not via PIM | extend Role = tostring(TargetResources[0].displayName) | extend AssignedUser = tostring(TargetResources[0].userPrincipalName) | extend AssignedBy = tostring(InitiatedBy.user.userPrincipalName) | project TimeGenerated, AssignedBy, AssignedUser, Role // Password Reset of Privileged User by Non-Admin let PrivilegedUsers = IdentityInfo | where AssignedRoles has_any ("Global", "Privileged", "Admin") | distinct AccountUPN; AuditLogs | where TimeGenerated > ago(7d) | where OperationName in ("Reset password", "Reset user password") | extend ResetUser = tostring(TargetResources[0].userPrincipalName) | extend ResetBy = tostring(InitiatedBy.user.userPrincipalName) | where ResetUser in (PrivilegedUsers) | where ResetBy !in (PrivilegedUsers) // Non-admin resetting admin | project TimeGenerated, ResetBy, ResetUser, OperationName

// Token Theft / Session Hijacking Indicators // Same User, Multiple IPs in Short Window SigninLogs | where TimeGenerated > ago(24h) | where ResultType == 0 | summarize IPCount = dcount(IPAddress), IPs = make_set(IPAddress), Locations = make_set(Location) by UserPrincipalName, bin(TimeGenerated, 15m) | where IPCount > 3 | project TimeGenerated, UserPrincipalName, IPCount, IPs, Locations // Sign-ins Without MFA After Successful MFA SigninLogs | where TimeGenerated > ago(24h) | where ResultType == 0 | extend MFAUsed = tostring(AuthenticationRequirement) | summarize MFASessions = countif(MFAUsed == "multiFactorAuthentication"), NoMFASessions = countif(MFAUsed == "singleFactorAuthentication") by UserPrincipalName | where MFASessions > 0 and NoMFASessions > 0 // Suspicious User Agent Strings SigninLogs | where TimeGenerated > ago(7d) | where ResultType == 0 | extend UA = tostring(DeviceDetail.browser) | where UA has_any ("curl", "python", "wget", "powershell", "httpie") | project TimeGenerated, UserPrincipalName, UA, IPAddress, AppDisplayName // Sign-ins to Sensitive Apps from New Device let SensitiveApps = dynamic(["Azure Portal", "Microsoft Azure", "Office 365 Exchange Online", "Microsoft Graph"]); let KnownDevices = SigninLogs | where TimeGenerated between(ago(30d) .. ago(1d)) | where ResultType == 0 | distinct UserPrincipalName, DeviceDetail; SigninLogs | where TimeGenerated > ago(1d) | where ResultType == 0 | where AppDisplayName in (SensitiveApps) | join kind=leftanti KnownDevices on UserPrincipalName, DeviceDetail | project TimeGenerated, UserPrincipalName, AppDisplayName, DeviceDetail

Using Bookmarks: When hunting query finds suspicious activity: 1. SELECT ROWS: - Check boxes next to suspicious entries 2. ADD BOOKMARK: Click "Add bookmark" - Name: "Suspicious OAuth app - Finance user" - Notes: Add investigation notes - Tags: "OAuth", "Persistence" - Entity mapping: Map user, IP, etc. 3. VIEW BOOKMARKS: Hunting Bookmarks - All saved findings - Can filter/search 4. CREATE INCIDENT FROM BOOKMARKS: Select bookmarks "Create incident" - Combines bookmarks into incident - Entities automatically mapped 5. SHARE WITH TEAM: - Bookmarks visible to team - Add to investigation notes Benefits: - Don't lose hunting findings - Build case over time - Easy incident creation - Collaboration on investigations