A baseline methodology for internal network engagements, from initial network foothold through domain compromise. Suitable as a starting point for most internal infrastructure or Active Directory-focused engagements.
Identify all live hosts, open ports, and running services across the target network range.
1. Perform host discovery across the assigned IP range/CIDR (ping sweep, ARP scan for local segments). 2. Run a full TCP port scan against all live hosts (Nmap -p- or masscan for speed). 3. Follow up with service/version detection on discovered open ports (Nmap -sV). 4. Identify Windows hosts, domain controllers, and other high-value targets by service fingerprint (LDAP, Kerberos, SMB). 5. Document the full host and service inventory for use in later phases.
Enumerate exposed Windows services for misconfigurations, shares, and version-specific vulnerabilities.
1. Enumerate SMB shares and permissions on all discovered hosts (smbclient, smbmap, crackmapexec). 2. Check for SMB signing disabled and known SMB vulnerabilities (EternalBlue-class, relay eligibility). 3. Enumerate LDAP anonymously where possible to pull domain, OU, and object information. 4. Check RDP and WinRM exposure and authentication requirements (NLA enabled/disabled). 5. Record all findings that indicate weak service configuration for exploitation later.
Attempt to extract domain information without valid credentials via null/guest sessions.
1. Attempt a null SMB session against each host (smbclient -N, rpcclient -U ""). 2. If successful, enumerate users, groups, and shares via the null session. 3. Attempt anonymous LDAP bind against the domain controller. 4. Check for SNMP with default/public community strings. 5. Document any information disclosed without authentication - this shapes the password spraying user list.
Capture NTLM hashes from broadcast name resolution poisoning, where in scope.
1. Confirm LLMNR/NBT-NS poisoning is in scope and approved before running (this is disruptive to normal traffic). 2. Run Responder (or equivalent) in analyze mode first to confirm broadcast traffic exists. 3. Run Responder in poisoning mode for an agreed time window to capture NTLMv2 hashes. 4. Attempt offline cracking of captured hashes with a strong wordlist/rules (Hashcat). 5. Document captured credentials and the hosts they originated from.
Test the harvested/generated username list against common passwords without triggering lockout.
1. Build a username list from prior enumeration (null sessions, OSINT, naming convention guesses). 2. Confirm the domain's account lockout threshold before spraying to avoid locking out real users. 3. Spray a small set of common/seasonal passwords across the username list with adequate delay between attempts. 4. Validate any successful credentials against a low-risk service (LDAP bind, SMB) before further use. 5. Document all valid credentials obtained and their associated privilege level.
Map the domain's trust and privilege relationships to identify attack paths to Domain Admin.
1. Collect AD data using a BloodHound collector (SharpHound or bloodhound-python) with the lowest-privilege valid account available. 2. Import the collected data into BloodHound and review default queries (Shortest Path to Domain Admins, Kerberoastable Users). 3. Identify high-value targets: Domain Admins, unconstrained delegation hosts, and ACL misconfigurations. 4. Map at least one credible attack path from the current foothold to Domain Admin. 5. Document the attack path graph as supporting evidence for the finding.
Extract and crack Kerberos service ticket hashes for accounts with weak or absent pre-authentication.
1. Enumerate service accounts with an SPN set (GetUserSPNs.py or PowerView) using any valid domain credential. 2. Request TGS tickets for each SPN account and extract the crackable hash (Kerberoasting). 3. Enumerate accounts with Kerberos pre-authentication disabled and request AS-REP hashes (GetNPUsers.py). 4. Attempt offline cracking of all extracted hashes with a strong wordlist/rules. 5. Document any cracked service account credentials and their privilege level.
Escalate from a standard user to local administrator on any host where a foothold was obtained.
1. Run an automated privilege escalation enumeration script (WinPEAS, PowerUp) on the compromised host. 2. Check for unquoted service paths, weak service permissions, and AlwaysInstallElevated misconfigurations. 3. Check for stored/cached credentials (Registry, unattended install files, scheduled tasks). 4. Confirm and exploit the most reliable escalation vector found to obtain local administrator. 5. Document the escalation technique and the misconfiguration that enabled it.
Move to additional hosts in the domain using captured credentials or tickets, without cracking passwords.
1. Attempt Pass-the-Hash against other hosts using any captured NTLM hash (crackmapexec, psexec.py). 2. Attempt Pass-the-Ticket using any captured Kerberos ticket where applicable. 3. Use WinRM/PsExec/WMI to obtain remote code execution on additional reachable hosts. 4. Repeat local enumeration and privilege escalation on each newly compromised host. 5. Maintain a running map of compromised hosts, credentials used, and access level obtained on each.
Execute the mapped attack path to obtain Domain Admin or equivalent domain-wide compromise.
1. Re-validate the BloodHound attack path against current compromised credentials/hosts. 2. Execute each step of the path (e.g. ACL abuse, unconstrained delegation abuse, GPO abuse) in order. 3. Confirm Domain Admin (or equivalent) access has been obtained via a low-risk validation action (e.g. reading a restricted share). 4. Avoid destructive actions (DCSync dumping full NTDS, golden ticket creation) unless explicitly authorized in scope. 5. Document the complete path from initial foothold to domain compromise, step by step.
Assess whether the environment would detect or prevent common persistence and evasion techniques (non-destructive validation only).
1. Review EDR/AV coverage on tested hosts and note any gaps observed during the engagement. 2. Test whether common living-off-the-land binaries were flagged or blocked during testing. 3. Note whether any alerting was observed (client SOC contact, if applicable) in response to test activity. 4. Do not deploy persistent backdoors or scheduled tasks unless explicitly scoped as a red team exercise. 5. Document detection/prevention gaps observed as findings for the blue team.
Consolidate the full compromise path with evidence, severity, and remediation guidance.
1. Compile all confirmed findings into a single report, ordered by the attack path taken. 2. Assign a severity to each finding based on impact (e.g. Domain Admin compromise = Critical) and exploitability. 3. Attach evidence (command output, screenshots) for each step of the compromise path. 4. Write remediation guidance specific to each finding (e.g. disable SMB signing exceptions, enforce LAPS, tier the AD model). 5. Review the report for false positives and confirm no out-of-scope hosts were touched before delivery.