H2buster Save

A threaded, recursive, web directory brute-force scanner over HTTP/2.

Project README

h2buster (v0.4d)

A fast, threaded, recursive, web directory brute-force scanner over HTTP/2 using hyper, inspired by Gobuster.

Features

  • Fast and portable - install hyper and run.
  • Multiconnection scanning.
  • Multithreaded connections.
  • Scalable: scans can be as docile or aggressive as you configure them to be.
  • h2 and h2c support.
  • Configurable directory recursion depth.
  • robots.txt scanning.
  • Multiplatform: works on both *nix and Windows.

Install

Requires Python 3.6. You only need to install one dependency. If you don't have hyper, run:
pip3 install -r requirements.txt

Usage

usage: h2buster.py [-h] -w wordlist -u target [-c connections=8]
                   [-t threads=20] [-m http_method=HEAD]
                   [-r directory_depth=2] [-hd header_list]
                   [-x extension_list] [-b http_code_list] [-l] [-vr] [-wc]
                   [-rb] [-nc]

h2buster: an HTTP/2 web directory brute-force scanner.

arguments:
  -h, --help            show this help message and exit
  -w wordlist           Directory wordlist
  -u target             Target URL/IP address ([scheme://]host[:port]).
                        Default port is 443 and HTTPS enabled. To specify
                        otherwise, use ':port' and/or 'http://' (port will
                        default to 80 then).
  -c connections=8      Number of HTTP/2 connections.
  -t threads=20         Number of threads per connection.
  -m http_method=HEAD   HTTP request method. Allowed values are GET, HEAD.
  -r directory_depth=2  Maximum recursive directory depth. Minimum is 1,
                        unlimited is 0.
  -hd header_list       List of headers in the format
                        'header->value|header->value...'. For example: -hd
                        'user-agent->Mozilla/5.0|accept-encoding->gzip,
                        deflate, br'.
  -x extension_list     List of file extensions to check separated by a
                        vertical bar (|). For example, -x '.php|.js|blank|/'.
                        The 'blank' keyword signifies no file extension.
                        Default extensions are '/', 'blank', '.html', '.php'
  -b http_code_list     List of blacklisted response codes separated by a
                        vertical bar (|). Directories with these response
                        codes will not be shown in the output. Default is 404.
  -l                    Flag: show response length in output. This overrides
                        the request method to GET.
  -vr                   Flag: force TLS certificate verification.
  -wc                   Flag: request a random path and analyze response to
                        detect false positives (wildcard processing).
  -rb                   Flag: scan for a robots.txt file. If found, a prompt
                        will be displayed asking whether to use the results.
  -nc                   Flag: disable colored output text.

Example

$ python3 h2buster.py -w test/small.txt -u www.google.com -c4 -t15 -m GET -x 'blank|/' -r1 -hd 'user-agent->h2buster' -b '404|301'

This will scan www.google.com, over 4 connections with 15 threads each, performing GET requests over the input wordlist, small.txt. Each entry in the wordlist will be sent as-is, as well as with a slash (/) appended to it. Found directories will not be scanned recursively (depth is 1). The string h2buster will be sent as an user agent. The response codes 404 and 301 will not be shown in the output.

Contributing

Check the TODO file for a list of features that need work.

Open Source Agenda is not affiliated with "H2buster" Project. README Source: 00xc/h2buster
Stars
36
Open Issues
0
Last Commit
4 years ago
Repository
License

Open Source Agenda Badge

Open Source Agenda Rating