external/anthropic-cybersecurity-skills/skills/deploying-active-directory-honeytokens/SKILL.md
Deploys deception-based honeytokens in Active Directory including fake privileged accounts with AdminCount=1, fake SPNs for Kerberoasting detection (honeyroasting), decoy GPOs with cpassword traps, and fake BloodHound paths. Monitors Windows Security Event IDs 4769, 4625, 4662, 5136 for honeytoken interaction. Use when implementing AD deception defenses for detecting lateral movement, credential theft, and reconnaissance.
npx skillsauth add seikaikyo/dash-skills deploying-active-directory-honeytokensInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Traditional signature-based detection misses novel attack techniques. Honeytokens provide high-fidelity detection with near-zero false positives because any interaction with a decoy object is inherently suspicious. In Active Directory:
| Event ID | Description | Honeytoken Use | |----------|-------------|----------------| | 4769 | Kerberos TGS ticket requested | Detect Kerberoast against honey SPN | | 4625 | Failed logon attempt | Detect use of fake credentials from decoy GPO | | 4662 | Directory service object accessed | Detect DACL read on honeytoken user | | 5136 | Directory service object modified | Detect modification of decoy GPO | | 5137 | Directory service object created | Detect GPO creation mimicking decoy | | 4768 | Kerberos TGT requested | Detect AS-REP roasting of honey account |
Per Trimarc Security research, effective honeytokens must appear legitimate:
Create a honeytoken account that mimics a legacy privileged service account.
# Import the deployment module
Import-Module .\scripts\Deploy-ADHoneytokens.ps1
# Create a honeytoken admin account
$honeyAdmin = New-HoneytokenAdmin `
-SamAccountName "svc_sqlbackup_legacy" `
-DisplayName "SQL Backup Service (Legacy)" `
-Description "Legacy SQL Server backup service account - DO NOT DELETE" `
-OU "OU=Service Accounts,DC=corp,DC=example,DC=com" `
-PasswordLength 128 `
-SetAdminCount $true
Write-Host "Honeytoken admin created: $($honeyAdmin.DistinguishedName)"
Assign a realistic but fake SPN to the honeytoken account. Any TGS request for this SPN is definitively malicious (honeyroasting).
# Add fake SPN to honeytoken account
$honeySPN = Add-HoneytokenSPN `
-SamAccountName "svc_sqlbackup_legacy" `
-ServiceClass "MSSQLSvc" `
-Hostname "sql-legacy-bak01.corp.example.com" `
-Port 1433
Write-Host "Honey SPN registered: $($honeySPN.SPN)"
Write-Host "Monitor Event ID 4769 for TGS requests targeting this SPN"
Create a fake GPO in SYSVOL with an embedded cpassword (Group Policy Preference password). Attackers using tools like Get-GPPPassword or gpp-decrypt will find and attempt to use these credentials, triggering detection.
# Create decoy GPO with cpassword trap
$decoyGPO = New-DecoyGPO `
-GPOName "Server Maintenance Policy (Legacy)" `
-DecoyUsername "admin_maintenance" `
-DecoyDomain "CORP" `
-SYSVOLPath "\\corp.example.com\SYSVOL\corp.example.com\Policies" `
-EnableAuditSACL $true
Write-Host "Decoy GPO created: $($decoyGPO.GPOGuid)"
Write-Host "SACL audit enabled - any read attempt will generate Event ID 4663"
Set ACL permissions that create fake attack paths visible to BloodHound/SharpHound reconnaissance, leading attackers toward monitored honeytokens.
# Create fake BloodHound attack path
$deceptivePath = New-DeceptiveBloodHoundPath `
-HoneytokenSamAccount "svc_sqlbackup_legacy" `
-TargetHighValueGroup "Domain Admins" `
-IntermediateOU "OU=Service Accounts,DC=corp,DC=example,DC=com"
Write-Host "Deceptive path created: $($deceptivePath.PathDescription)"
Set up SIEM detection rules to alert on any honeytoken interaction.
# Using the Python detection agent
from agent import ADHoneytokenMonitor
monitor = ADHoneytokenMonitor(config_path="honeytoken_config.json")
# Register all honeytokens for monitoring
monitor.register_honeytoken("svc_sqlbackup_legacy", token_type="admin_account")
monitor.register_honeytoken("MSSQLSvc/sql-legacy-bak01.corp.example.com:1433", token_type="spn")
monitor.register_honeytoken("admin_maintenance", token_type="gpo_credential")
# Generate SIEM detection rules
splunk_rules = monitor.generate_detection_rules(siem="splunk")
sentinel_rules = monitor.generate_detection_rules(siem="sentinel")
sigma_rules = monitor.generate_detection_rules(siem="sigma")
for rule in sigma_rules:
print(f"Rule: {rule['title']}")
print(f" Detection: {rule['detection_logic']}")
Test the honeytokens to ensure detection fires correctly.
# Validate honeytoken deployment
$validation = Test-HoneytokenDeployment `
-SamAccountName "svc_sqlbackup_legacy" `
-ValidateAdminCount `
-ValidateSPN `
-ValidateGPODecoy `
-ValidateAuditPolicy
$validation | Format-Table Check, Status, Details -AutoSize
Import-Module .\scripts\Deploy-ADHoneytokens.ps1
# Deploy complete honeytoken suite
$deployment = Deploy-FullHoneytokenSuite `
-Environment "Production" `
-ServiceAccountOU "OU=Service Accounts,DC=corp,DC=example,DC=com" `
-SYSVOLPath "\\corp.example.com\SYSVOL\corp.example.com\Policies" `
-TokenCount 3 `
-IncludeSPN $true `
-IncludeGPODecoy $true `
-IncludeBloodHoundPath $true `
-SIEMType "Splunk"
# Output deployment report
$deployment.Tokens | Format-Table Name, Type, SPN, DetectionRule -AutoSize
$deployment | Export-Csv "honeytoken_deployment_report.csv" -NoTypeInformation
index=wineventlog EventCode=4769 ServiceName="svc_sqlbackup_legacy"
| eval alert_severity="critical"
| eval alert_type="honeytoken_kerberoast"
| table _time, src_ip, Account_Name, ServiceName, Ticket_Encryption_Type
| sort - _time
SecurityEvent
| where EventID == 4769
| where ServiceName in ("svc_sqlbackup_legacy", "svc_exchange_legacy")
| extend AlertType = "Honeytoken Kerberoast Detected"
| project TimeGenerated, Computer, Account, ServiceName, IpAddress, TicketEncryptionType
development
拋棄式 HTML mockup 比稿:產出 2 到 3 個設計立場不同的變體(密度 / 版式 / 強調軸,不是換色),各附取捨說明,最後給有立場的對比結論。適用:「畫個草圖」「比較 A 版 B 版」「先看方向再做」「給我看幾種做法」。要 production 元件或設計已定案時不適用。
tools
需求不明時的意圖萃取訪談:一次一題、每題附上自己的猜測、聽出「真正想要 vs 覺得應該要」,直到能預測使用者反應(約 95% 信心)才動工。適用:需求缺少對象 / 動機 / 成功標準 / 約束,或使用者點名「訪談我」「先確認一下」「我們確定嗎」。明確自足的指示、純資訊查詢、機械性操作不適用。
development
對非平凡決策啟動新鮮 context 對抗審查(找碴不背書),在修正還便宜的時候抓出錯誤方向。適用:高風險改動(production、資安敏感邏輯、不可逆操作)、不熟的程式碼、要宣稱「這樣是安全的 / 可行的」之前。機械性操作與一行修改不適用。
testing
Reference for writing and editing agent skills well — the vocabulary and principles that make a skill predictable. Consult when authoring, reviewing, or pruning a SKILL.md.