NoXss is a cross-site script vulnerability scanner supported reflected xss and dom-based xss. It's very fast and suitable for testing millions of urls. It has found some xss vulnerabilities in Bug Bounty program.
- Fast and suitable for testing millions of urls
- Support Dom-based xss(use Chrome or Phantomjs) and reflected xss
- Only use 8 Payloads based on injection postion now(not fuzz,more accurate,faster)
- Async request(use gevent) and Multi-processed
- Support single url,file and traffic from Burpsuite
- Traffic filter based on interface
- Support speicial headers(referer,cookie,customized token,e.g.)
- Support rescan quickly by id
├── engine.py
├── logo
├── cookie.py
├── url.txt
├── cookie
│ └── test.com_cookie
├── traffic
│ ├── 49226b2cbc77b71b.traffic #traffic file(pickled)
│ └── 49226b2cbc77b71b.reflect #reflected file(pickled)
├── config.py
├── start.py
├── url.txt.filtered #filtered urls
├── util.py
├── README.md
├── banner.py
├── requirements.txt
├── result
│ └── 49226b2cbc77b71b-2019_10_29_11_24_44.json #result
├── model.py
└── test.py
Linux
Python2.7
Browser:Phantomjs or Chrome
- 1.
apt-get install flex bison phantomjs - 2.
pip install -r requirements.txt
- 1.
yum install flex bison phantomjs - 2.
pip install -r requirements.txt
- 1.
brew install grep findutils flex phantomjs - 2.
pip install -r requirements.txt
If you want to scan use "--browser=chrome",you must install chrome mannually. You can use "--check" to test the installation.
python start.py --check
python start.py --url url --save
python start.py --url url --cookie cookie --browser chrome --save
python start.py --url url --cookie cookie --browser chrome-headless --save
python start.py --file ./url.txt --save
python start.py --burp ./test.xml --save
python start.py --file file --filter
--url scan from url.
--id rescan from *.traffic file by task id.
--file scan urls from text file(like ./url.txt).
--burp scan *.xml(base64 encoded,like ./test.xml) from burpsuite proxy.
--process number of process.
--coroutine number of coroutine.
--cookie use cookie.
--filter filter urls.
--browser use browser(chrome,chrome-headless or phantomjs) to scan,it's good at DOM-based xss but slow.
--save save results to ./result/id.json.
--clear delete traffic files after scanning.
In Proxy,"Save items" ==> "test.xml"
Then you can scan test.xml:
python start.py --burp=./test.xml
After scanning firstly,there will be taskid.traffic and taskid.reflect in ./traffic/:
- taskid.traffic: Web traffic of request(pickled).
- taskid.reflect: Reflected result (pickled)that included reflected params,reflected position,type and others.
NoXss will use these middle files to rescan:
python start.py --id taskid --save
NoXss use only 8 payloads for scanning.These payloads are based on param's reflected position.Fewer payloads make it faster than fuzzing.
NoXss is highly concurrent for using coroutine.
More and more page is using dom to render html.NoXss can parse it with using Phantomjs(default) or chrome.
Some xss is difficult to scan.NoXss will save some files in traffic/ for analysing,include:
- *.traffic(traffic file during scanning)
- *.reflect(param's reflected result)
- *.redirect(30x response)
- *.error(some error happened such as timeout,connection reset,etc.)
- *.multipart(when request is multupart-formed,not easy to scan)
As you see in Screenshot,the poc is https://xxx/?proxyAccount=xssjs%22%3B&shareName=duhxams,That means use the payload xssjs%22%3B in param "proxyAccount":
Then you can end the double qoutes use payload xssjs";alert(1);//.The final exploit is:
https://xxx.com/?proxyAccount=xssjs";alert(1);//&shareName=duhxams