Post

TryHackMe — VulnNet-RST | Active Directory attack-chain assessment

Evidence-based case study of a VulnNet-RST Windows domain lab: SMB and account enumeration, AS-REP Roasting, WinRM access and DRSUAPI credential extraction.

TryHackMe — VulnNet-RST | Active Directory attack-chain assessment

Platform: TryHackMe · Lab: VulnNet-RST · Environment: Windows Server / Active Directory · Work performed: April 2026 · Type: Authorized training lab

Executive summary

This case study documents my hands-on assessment of the VulnNet-RST Active Directory environment. The exercise connected several stages of an internal Windows domain attack: discovering domain users, identifying an account exposed to AS-REP Roasting, recovering a password offline, assessing SMB access, obtaining an authorized lab shell over WinRM, and validating the impact of directory replication rights through DRSUAPI-based credential extraction.

The strongest verified result was the successful retrieval of Administrator NTLM material and Kerberos keys using Impacket’s secretsdump functionality. This is evidence of a serious Active Directory privilege boundary failure within the lab. It is not a claim of a production engagement, and the records available to me do not establish that I captured a final room flag.

This write-up now includes commands from my April 2026 terminal session, with short interpretations of the results. Where the exact historical command was not preserved, I say so instead of presenting a guessed command as something I ran. Passwords, hashes and flags are omitted; commands that originally embedded a password are shown in an equivalent sanitized form using shell variables.

Lab IP note: TryHackMe target addresses changed between sessions. My initial enumeration targeted 10.128.134.155; later authenticated testing targeted 10.128.182.129. These were temporary lab addresses, not permanent infrastructure.

1. Scope and objectives

The target was a Windows Server domain environment with common enterprise authentication and file-sharing services. The practical objectives were to:

  1. Discover exposed Windows and Active Directory services.
  2. Enumerate domain identities and accessible SMB resources.
  3. Assess Kerberos pre-authentication exposure and password strength.
  4. Test whether recovered credentials enabled remote management access.
  5. Determine the effective impact of privileged directory permissions.
  6. Document attack-path causes and defensive measures.

All actions were performed within the training environment.

2. Validated attack-chain overview

The following diagram summarizes the observed phases. It intentionally does not invent the exact intermediate mechanism by which the second account’s credentials were discovered.

flowchart TD
    A["Windows / AD service discovery"] --> B["SMB and domain user enumeration"]
    B --> C["AS-REP Roasting<br/>Offline password recovery"]
    C --> D["SMB access and share review"]
    D --> E["Further credential discovery<br/>Secondary account"]
    E --> F["Authenticated WinRM shell"]
    F --> G["DRSUAPI directory replication<br/>Impacket secretsdump"]
    G --> H["Administrator NTLM material<br/>and Kerberos keys retrieved"]

Important distinction: the first recovered account did not immediately provide WinRM access. The later WinRM shell used a different account. The exact source of that second account’s credentials is not reconstructed here because the retained evidence is insufficient to describe it reliably.

3. Reconnaissance and SMB enumeration

The initial work focused on the domain’s Windows services and the information available through SMB and account enumeration. Guest/anonymous access was useful for establishing the environment’s structure, and RID-based enumeration helped identify candidate domain accounts.

With a recovered domain user’s credentials, I was able to authenticate to SMB and review accessible shares. The documented environment exposed standard domain resources such as NETLOGON and SYSVOL, along with additional business-oriented shares.

This stage illustrated two points:

  • An account with limited permissions can still reveal information that is useful for further assessment.
  • A readable share is not automatically a vulnerability; the security impact depends on which information is exposed and whether it enables a meaningful next step.

Terminal evidence: domain enumeration

The following command is preserved from my initial terminal session:

1
enum4linux -A 10.128.134.155

The output identified the VULNNET-RST domain, permitted an anonymous/null SMB session, and exposed domain identity information including the SID:

1
2
Domain Name: VULNNET-RST
Domain SID: S-1-5-21-1589833671-435344116-4136949213

RID enumeration was part of the account-discovery workflow, but I do not have its exact historical command line preserved.

Terminal evidence: authenticated SMB

After the lab IP changed, I validated the recovered t-skid account with NetExec. The original password has been removed from the command below.

1
2
read -rsp 'Password for t-skid: ' TSKID_PASSWORD; echo
nxc smb 10.128.182.129 -u 't-skid' -p "$TSKID_PASSWORD" --shares

The recorded output identified Windows Server 2019 (build 17763) in vulnnet-rst.local, confirmed successful authentication and listed readable shares, including `IPC— layout: post title: “TryHackMe — VulnNet-RST | Active Directory attack-chain assessment” date: 2026-11-10 00:00:00 +0100 categories: [TryHackMe, Active Directory] tags: [tryhackme, active-directory, kerberos, smb, asrep-roasting, evil-winrm, impacket, dcsync, pentest] description: “Evidence-based case study of a VulnNet-RST Windows domain lab: SMB and account enumeration, AS-REP Roasting, WinRM access and DRSUAPI credential extraction.” mermaid: true —

Platform: TryHackMe · Lab: VulnNet-RST · Environment: Windows Server / Active Directory · Work performed: April 2026 · Type: Authorized training lab

Executive summary

This case study documents my hands-on assessment of the VulnNet-RST Active Directory environment. The exercise connected several stages of an internal Windows domain attack: discovering domain users, identifying an account exposed to AS-REP Roasting, recovering a password offline, assessing SMB access, obtaining an authorized lab shell over WinRM, and validating the impact of directory replication rights through DRSUAPI-based credential extraction.

The strongest verified result was the successful retrieval of Administrator NTLM material and Kerberos keys using Impacket’s secretsdump functionality. This is evidence of a serious Active Directory privilege boundary failure within the lab. It is not a claim of a production engagement, and the records available to me do not establish that I captured a final room flag.

This write-up now includes commands from my April 2026 terminal session, with short interpretations of the results. Where the exact historical command was not preserved, I say so instead of presenting a guessed command as something I ran. Passwords, hashes and flags are omitted; commands that originally embedded a password are shown in an equivalent sanitized form using shell variables.

Lab IP note: TryHackMe target addresses changed between sessions. My initial enumeration targeted 10.128.134.155; later authenticated testing targeted 10.128.182.129. These were temporary lab addresses, not permanent infrastructure.

1. Scope and objectives

The target was a Windows Server domain environment with common enterprise authentication and file-sharing services. The practical objectives were to:

  1. Discover exposed Windows and Active Directory services.
  2. Enumerate domain identities and accessible SMB resources.
  3. Assess Kerberos pre-authentication exposure and password strength.
  4. Test whether recovered credentials enabled remote management access.
  5. Determine the effective impact of privileged directory permissions.
  6. Document attack-path causes and defensive measures.

All actions were performed within the training environment.

2. Validated attack-chain overview

The following diagram summarizes the observed phases. It intentionally does not invent the exact intermediate mechanism by which the second account’s credentials were discovered.

flowchart TD
    A["Windows / AD service discovery"] --> B["SMB and domain user enumeration"]
    B --> C["AS-REP Roasting<br/>Offline password recovery"]
    C --> D["SMB access and share review"]
    D --> E["Further credential discovery<br/>Secondary account"]
    E --> F["Authenticated WinRM shell"]
    F --> G["DRSUAPI directory replication<br/>Impacket secretsdump"]
    G --> H["Administrator NTLM material<br/>and Kerberos keys retrieved"]

Important distinction: the first recovered account did not immediately provide WinRM access. The later WinRM shell used a different account. The exact source of that second account’s credentials is not reconstructed here because the retained evidence is insufficient to describe it reliably.

3. Reconnaissance and SMB enumeration

The initial work focused on the domain’s Windows services and the information available through SMB and account enumeration. Guest/anonymous access was useful for establishing the environment’s structure, and RID-based enumeration helped identify candidate domain accounts.

With a recovered domain user’s credentials, I was able to authenticate to SMB and review accessible shares. The documented environment exposed standard domain resources such as NETLOGON and SYSVOL, along with additional business-oriented shares.

This stage illustrated two points:

  • An account with limited permissions can still reveal information that is useful for further assessment.
  • A readable share is not automatically a vulnerability; the security impact depends on which information is exposed and whether it enables a meaningful next step.

, NETLOGON, SYSVOL, VulnNet-Business-Anonymous and VulnNet-Enterprise-Anonymous.

I then listed shares and browsed a business share using smbclient (password passed through the same shell variable rather than embedded in the public write-up):

1
2
smbclient -L //10.128.182.129/ -U "t-skid%$TSKID_PASSWORD"
smbclient //10.128.182.129/VulnNet-Business-Anonymous -U "t-skid%$TSKID_PASSWORD"

Inside the interactive SMB session:

1
2
3
4
smb: \\> ls
  Business-Manager.txt
  Business-Sections.txt
  Business-Tracking.txt

Interpretation: valid low-privilege domain credentials gave access to multiple shares. The listing alone does not prove sensitive data exposure; that requires reviewing actual file contents and permissions.

Relevant tools/protocols: enum4linux, SMB, NetExec, smbclient, Windows domain account enumeration.

4. Kerberos exposure: AS-REP Roasting

The account-enumeration stage identified an identity susceptible to AS-REP Roasting. When Kerberos pre-authentication is not required for an account, an attacker may request authentication material that can be tested offline against candidate passwords.

In this lab, the AS-REP material was successfully cracked and a working domain password was recovered. I then validated the account’s access, including SMB authentication.

Terminal evidence: offline password recovery

I saved the AS-REP response in hash.txt and ran the following recorded John the Ripper command:

1
john --wordlist=/usr/share/wordlists/rockyou.txt hash.txt

John recovered the password for the t-skid account. The recovered secret is not published here. The precise historical command used to request the AS-REP material was not preserved, so I have not invented one.

Terminal evidence: WinRM access denied

I then attempted WinRM with t-skid:

1
evil-winrm -i 10.128.182.129 -u t-skid -p "$TSKID_PASSWORD"

The attempt produced a WinRM authorization error; it did not establish a usable administrative shell. This was an important finding: valid domain credentials do not imply permission to log in remotely.

Defensive interpretation

  • Require Kerberos pre-authentication for accounts unless there is a justified exception.
  • Enforce strong, unique passwords and review service-account configurations.
  • Monitor anomalous Kerberos authentication requests, while correlating them with account configuration and other events.
  • Restrict remote management rights independently of basic domain authentication.

5. WinRM access and privilege assessment

After further investigation, a second domain account was used to obtain an authenticated Windows shell through WinRM. This allowed the lab’s domain permission model to be assessed from an authenticated context.

I did not treat shell access itself as proof of domain administration. Instead, the next step was to verify the effective directory permissions by testing a concrete capability.

Terminal evidence: successful shell with another account

The later authenticated session used a different domain identity, a-whitehat. To avoid publishing the lab password, this version prompts for it:

1
2
read -rsp 'Password for a-whitehat: ' AWHITEHAT_PASSWORD; echo
evil-winrm -i 10.128.182.129 -u a-whitehat -p "$AWHITEHAT_PASSWORD"

Recorded shell validation:

1
2
PS C:\\Users\\a-whitehat\\Documents> whoami
vulnnet-rst\\a-whitehat

The evidence confirms a working WinRM shell as a-whitehat. It does not establish precisely how this account’s credentials were found, so that intermediate step is left unclaimed.

Relevant tools/protocols: Evil-WinRM, Windows authentication, Active Directory permission analysis.

6. DRSUAPI credential extraction: verified impact

The decisive result came from an Impacket secretsdump operation targeting the Administrator account via DRSUAPI. The recorded command completed successfully and returned:

  • NTLM credential material for the domain Administrator account.
  • Associated Kerberos AES key material.

Terminal evidence: targeted directory replication

The following command preserves the tool, account, target and options I actually used, but substitutes the previously prompted password for the original inline secret:

1
impacket-secretsdump "vulnnet-rst.local/a-whitehat:$AWHITEHAT_PASSWORD@10.128.182.129" -just-dc-user Administrator

The observed result, redacted for publication, had this structure:

1
2
3
4
5
[*] Using the DRSUAPI method to get NTDS.DIT secrets
vulnnet-rst.local\\Administrator:500:aad3b435b51404eeaad3b435b51404ee:[NTLM_REDACTED]:::
[*] Kerberos keys grabbed
Administrator:aes256-cts-hmac-sha1-96:[AES256_REDACTED]
Administrator:aes128-cts-hmac-sha1-96:[AES128_REDACTED]

This is a redacted schematic of the observed result, not a byte-for-byte terminal transcript. The crucial verified fact is that the operation returned the Administrator’s NTLM material and Kerberos keys.

This demonstrates the impact of directory replication privileges: an identity able to perform sensitive replication operations may extract password-related material from Active Directory without needing interactive login as the affected user.

The key technical lesson is to distinguish being able to authenticate, having remote shell rights, and possessing directory replication privileges. These are separate security boundaries, and the last can be particularly sensitive.

Defensive interpretation

AreaRisk illustratedRecommended control
Kerberos accountsOffline password attacks against accounts without pre-authenticationAudit pre-authentication settings and password quality
SMB sharesInformation exposure to low-privilege identitiesApply least privilege and review sensitive file access
Remote managementUnauthorized interactive access when permissions are excessiveRestrict WinRM access to approved administrative identities
AD replicationExtraction of high-value domain credentialsAudit replication rights and privileged group memberships
DetectionMulti-stage attacks may appear as unrelated eventsCorrelate account enumeration, Kerberos requests, remote logons and replication activity

7. Evidence, limitations and outcome

Confirmed from retained lab evidence:

  • Domain account discovery and an AS-REP Roasting/password recovery workflow.
  • Successful SMB authentication and access to several domain shares.
  • An initial WinRM access denial with the first recovered account.
  • Subsequent WinRM shell access using another account.
  • Successful DRSUAPI extraction of Administrator NTLM material and Kerberos keys.

Not claimed: a verified final flag submission, a production-system compromise, the exact AS-REP request or RID-brute command line, or an exact intermediate credential-discovery technique where the evidence is incomplete.

Command provenance: enum4linux -A, john --wordlist, nxc smb --shares, the two smbclient sessions, both evil-winrm attempts and impacket-secretsdump -just-dc-user Administrator are grounded in the retained April 2026 lab session. Password handling and selected output have been sanitized for public publication.

8. What this project demonstrates

This lab gave me practical experience in Windows/AD enumeration, Kerberos security, SMB access assessment, credential validation, WinRM access, directory replication risks and evidence-based reporting.

For an offensive-security or Red Team internship, the value is not merely running tools: it is following the relationship between identity configuration, authentication boundaries, delegated rights and the actual security impact.

Authorized lab exercise. No passwords, hashes or target-specific secrets are published in this case study. Commands are shown only for this permitted training environment.

This post is licensed under CC BY 4.0 by the author.