fuzzchromium / Kite

Browser Fuzzing Framework ,浏览器Fuzzing 框架..

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

##New Architecture

Kite_Fuzzing_Architecture

Now developing

##How to using it

###Fuzzing 1.run open_console_in_this_dir.bat
1.运行open_console_in_this_dir.bat

step_1

2.run web_server.py in this new console window
2.因为web_server.py 需要当前目录定位到browser_fuzzing 中,所以使用open_console_in_this_dir.bat 或者直接双击,那么web_server.py 就会在browser_fuzzing 的目录中运行,否则会产生错误(找不到fuzzer 文件夹)

step_2

3.using your browser which you want to fuzzing to open http://127.0.0.1/vector
3.选择你希望Fuzzing 的浏览器来访问http://127.0.0.1/vector ,然后页面会自己刷新

step_3

4.using get_poc.py dump crash poc when browser has been crash
4.如果浏览器产生了崩溃,这个时候web_server.py 会保存刚才的页面缓存,然后使用get_poc.py 来获取崩溃页面的代码

step_4_1

step_4_2

step_4_3

5.repeat crash in browser
5.把崩溃文件拖放到原来的浏览器中重现崩溃

step_5_1

step_5_2

6.all auto fuzzing work include the process_monitor.py,you just config the browser path what you want to fuzzing and run it
6.以上就是Fuzzing 的基本流程,所有的工作都集合到了process_monitor.py ,只需要运行它就可以自动化Fuzzing 浏览器,崩溃的文件保存在poc 文件夹下


###Valid PoC File (验证崩溃文件的有效性) 1.run valid_server.py
1.运行valid_server.py ,valid_server.pyweb_server.py 的不同之处在于前者用于崩溃文件验证,后者用来生成崩溃文件,下面会解释为什么要这样设计

step_6

2.run valid_poc.py
2.运行valid_poc.py 来验证崩溃文件

step_7

WARNING! Sometime Browser Will Follwing Down in DeadLoop .Just Using taskmgr.exe to Kill it .valid_poc.py Will Recall Browser EXE
WARNING! 因为浏览器的设计原理不允许在file:// 域下直接执行脚本代码(浏览器的安全设计,防止XHR 文件探测和遍历等,而且file:// 域下的的HTTP 访问不带referer ),所以需要使用web 服务器来发送这些崩溃文件到浏览器执行
TIPS :If you get a Exploit File ,Congratulation you !This file will be save to dir exploit
TIPS :有价值的崩溃文件会保存到文件夹exploit

###How to using valid_poc.py (怎么使用valid_poc.py) valid_poc.py will help you dig valueable PoC in all crash files,using :

valid_poc.py count -- get all poc files number (获取崩溃文件数)
valid_poc.py %poc_index% -- test a poc (测试和验证崩溃文件)
valid_poc.py %poc_index% debug -- test a poc and output more detail (调试崩溃文件)

if you want to get more detail in analays crash files information ,plaese start valid_poc_debug.py.you can see that instruction where will crash
在使用valid_poc.py 的验证过程中可能因为速度太快没有办法看到具体是哪段代码哪些异常产生了崩溃,valid_poc_debug.py 解决了这个问题,只需要在准备验证崩溃文件之前启动valid_poc_debug.py ,它会自动帮你运行valid_server.pyvalid_poc.py ,然后在本地绑定端口,接收valid_poc.py 发过来的崩溃数据

step_8

###About valid_server.py You can not execute a JavaScript or VBScript in Internet Explorer still load in file:// .So I make this WEB Server to resolve this probleam ,this is Using : valid_server.py -- run server(it will load crash file from path \poc)
valid_server.py debug -- run server(it will load crash file from path \exploit)
不加选项用来验证所有的fuzzing 出来的文件(加载\poc 目录下的文件),debug 的选项是给后面调试有效的利用文件使用的(加载\exploit 目录下的文件)
if you want to debug a exploit file ,using valid_server.py debug and valid_poc.py %CrashFileIndex% debug .valid_poc.py include a small debugger (support:-r %regesit% (look regesit) ;-a %address% (look memory address) ;-u %address% (get instruction) ;-quit (will exit)).
为了方便调试(因为开Windbg 有时候比较繁琐),里面集合了一个精简的调试器,方便快速定位问题(比如空指针,栈溢出等),支持的命令有:-r %寄存器名字% (读取某个寄存器的数据),-a %内存地址% (读取内存数据),-u %内存地址% (反编译指定地址),-quit (退出调试)

if you want to all automation ,plaese using valid_poc_debug.py

###Distributed Fuzzing -- distributed_master.py and distributed_slave.py (分布式fuzzing 部分) Actually ,one machine can not fuzzing more browser either excellent performance.You can let it make more virtual machine to fuzzing .
单个操作系统是很难可以同时来fuzzing 多个浏览器的,通常情况下一个系统只能开一个fuzzing ,所以就需要开虚拟机来最大化地利用单台主机,但是有时候操作多台主机不方便,所以需要使用分布式的模块来方便监视和调控整个fuzzing 机组系统

distributed_master.py -- all slave fuzzing machine manager (master 管理整个fuzzing 系统),support command :
-update update new code to slave machine from master (更新框架代码)
-upload collect all slave machine fuzzing PoC file (调控fuzzing 系统上传崩溃文件)

distributed_slave.py -- run in slave fuzzing machine (fuzzing 机器上运行的控制模块)

step_9-1
step_9-2

###Other Envirment Setting (其他环境设置) 1.Set IE Single Process -> set_ie_single_process.reg
1.设置IE 为单进程

2.Set disable system crash tips window -> clear_system_debug_tips_window.reg
2.取消系统崩溃提示(有时候这些崩溃调试会阻塞接下来继续fuzzing 的操作.注意,里面需要设置调试进程为browser_fuzzing 目录下的disable_system_debug_tips_window.exe ,这个路径需要手动设置)


so ,this is a full step to collect a crash PoC ,but i write a auto fuzzing script -- process_monitor.py .You just need set some argerment about target fuzzing browser path .etc .Good luck ! ..
TIPS 1:If you have not install pydasm ,download link :https://github.com/reider-roque/pydbg-pydasm-paimei;If you not install psutil ,download link :https://pypi.python.org/pypi/psutil
TIPS 2:You need manual setting some setting about Internet Explorer .I packet these import setting to Regesit File ,care clear_system_debug_tips_window.reg(You Need Manual Setting Regesit Item Value to the diable_system_debug_tips_window.exe's full path ) and set_ie_single_process.reg
TIPS 3:Add valid_poc.py and valid_server.py .First ,start valid_server.py to run a WEB server for valid Fuzzing PoC is or not a valid Vuln ;and than ,run valid_poc.py call Browser getting this PoC to process and valid_poc.py will analysis crash file ,if this is a Vuln Crash File,it will save this PoC to dir exploit
TIPS 4:May you need to setting what browser you want to test path in process_monitor.py and valid_poc.py
TIPS 4:在process_monitor.pyvalid_poc.py 中设置浏览器路径

About

Browser Fuzzing Framework ,浏览器Fuzzing 框架..


Languages

Language:HTML 96.0%Language:Python 4.0%Language:Batchfile 0.0%