2024年---第十五届蓝桥杯网络安全CTF赛道Writeup(Web、Misc、Crypto、Reverse)
第一题:爬虫协议
1.1 题目描述:
小兰同学在开发网站时了解到一个爬虫协议,该协议指网站可建立一个特别的txt文件来告诉搜索引擎哪些页面可以抓取,哪些不可以抓取,而搜索引擎则通过读取该txt文件来识别这个页面是否允许被抓取,爬虫协议并不是一个规范而只是约定俗成的,所以不能保证网站的隐私。
1.2 题目界面:
打开题目网址
1.3 解题步骤:
1):尝试访问robots.txt文件
2):尝试访问0f3f8d7ec56c0c50f9d73724902648a5
3):尝试访问0f3f8d7ec56c0c50f9d73724902648a5/da87bd9aca438bf00a80d12c8912f3a0
4):得到flag{42d5c8f1-259d-462c-a8bd-4bc1f50d354d}
第二题:流量分析
2.1 解题步骤:
1):用wireshark打开题目所给文件
2):导出对象
3):选择HTTP对象导出
4):用记事本打开flag%20%7cbase64%20-w%200%27);文件
5):base64解密文件内容ZmxhZ3s3ZDZmMTdhNC0yYjBhLTQ2N2QtOGE0Mi02Njc1MDM2OGMyNDl9Cg==得到flag{7d6f17a4-2b0a-467d-8a42-66750368c249}
第三题:AES加密
3.1 题目界面:
3.2 解题步骤:
1):分析得出:题目界面已经告诉我们采用的是AES加密,并且告诉我们key是gamelab@gamelab@,IV是gamelab@gamelab@,Mode是CBC模式,输出为Hex十六进制4da72144967f1c25e6273950bf29342aae635e2396ae17c80b1bff68d90f16679bb45c15852e0ce88d4864d93e9e3be2
2):打开解密工具cc(CyberChef),输入key,IV,Mode以及密文,点击decode解密得到flag{6500e76e-15fb-42e8-8f29-a309ab73ba38}
第四题:RSA加密
4.1 题目文件为task.py,文件代码内容如下所示:
from Crypto.Util.number import * from gmpy2 import * flag = b'xxx' m = bytes_to_long(flag) p = getPrime(512) q = next_prime(p) e = 65537 n = p * q phi = (p - 1) * (q - 1) d = inverse(e, phi) d1 = d % q d2 = d % p c = pow(m, e, n) print(n) print(d1) print(d2) print(c) # 94581028682900113123648734937784634645486813867065294159875516514520556881461611966096883566806571691879115766917833117123695776131443081658364855087575006641022211136751071900710589699171982563753011439999297865781908255529833932820965169382130385236359802696280004495552191520878864368741633686036192501791 # 4218387668018915625720266396593862419917073471510522718205354605765842130260156168132376152403329034145938741283222306099114824746204800218811277063324566 # 9600627113582853774131075212313403348273644858279673841760714353580493485117716382652419880115319186763984899736188607228846934836782353387850747253170850 # 36423517465893675519815622861961872192784685202298519340922692662559402449554596309518386263035128551037586034375613936036935256444185038640625700728791201299960866688949056632874866621825012134973285965672502404517179243752689740766636653543223559495428281042737266438408338914031484466542505299050233075829
4.2 解题步骤:
1):分析得出:我们得知道这个算法的原理,就是现在p,q是两个素数Q,而且他俩在素数序列里面就是一前一后的关系。所以我们要把他俩的乘积开根号得到的结果一定是在p,q之间的一个数字,(而且一定不是素数,因为p,q就是紧邻的两个素数)。那我们找这个开方出来的数字的下一个素数,一定是q,因此我们再让n/q就可以得到两个素数。
2):解法一,得到flag{5f00e1b9-2933-42ad-b4e1-069f6aa98e9a}:
import gmpy2 import sympy # import Cryptodome.Util.number from Crypto.Util.number import * import binascii n=94581028682900113123648734937784634645486813867065294159875516514520556881461611966096883566806571691879115766917833117123695776131443081658364855087575006641022211136751071900710589699171982563753011439999297865781908255529833932820965169382130385236359802696280004495552191520878864368741633686036192501791 x=gmpy2.iroot(n,2)[0]# 取第0个元素,也就是第一个元素 p=sympy.nextprime(x) q=n//p e=65537 d=gmpy2.invert(e,(p-1)*(q-1)) print(p) print(q) print(d) c=36423517465893675519815622861961872192784685202298519340922692662559402449554596309518386263035128551037586034375613936036935256444185038640625700728791201299960866688949056632874866621825012134973285965672502404517179243752689740766636653543223559495428281042737266438408338914031484466542505299050233075829 m=pow(c,d,n) print(m) print(long_to_bytes(m))
3):解法二,直接yafu分解n,得到p,q的值,然后计算得到flag。
第五题:DWT盲水印
5.1 题目文件如下所示:
5.2 解题步骤:
1):破解orign压缩包:打开压缩包发现压缩包里面的serect.txt是密码字典,未加密,可以解压出来,然后用serect.txt字典破解压缩包orign.zip,得到a.png也就是原始图片。
2):打开lose.py分析发现为DWT盲水印代码:
class WaterMarkDWT: def __init__(self, origin: str, watermark: str, key: int, weight: list): self.key = key self.img = cv2.imread(origin) self.mark = cv2.imread(watermark) self.coef = weight def arnold(self, img): r, c = img.shape p = np.zeros((r, c), np.uint8) a, b = 1, 1 for k in range(self.key): for i in range(r): for j in range(c): x = (i + b * j) % r y = (a * i + (a * b + 1) * j) % c p[x, y] = img[i, j] return p def deArnold(self, img): r, c = img.shape p = np.zeros((r, c), np.uint8) a, b = 1, 1 for k in range(self.key): for i in range(r): for j in range(c): x = ((a * b + 1) * i - b * j) % r y = (-a * i + j) % c p[x, y] = img[i, j] return p def get(self, size: tuple = (1200, 1200), flag: int = None): img = cv2.resize(self.img, size) img1 = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY) img2 = cv2.cvtColor(self.mark, cv2.COLOR_RGB2GRAY) c = pywt.wavedec2(img2, 'db2', level=3) [cl, (cH3, cV3, cD3), (cH2, cV2, cD2), (cH1, cV1, cD1)] = c d = pywt.wavedec2(img1, 'db2', level=3) [dl, (dH3, dV3, dD3), (dH2, dV2, dD2), (dH1, dV1, dD1)] = d a1, a2, a3, a4 = self.coef ca1 = (cl - dl) * a1 ch1 = (cH3 - dH3) * a2 cv1 = (cV3 - dV3) * a3 cd1 = (cD3 - dD3) * a4 waterImg = pywt.waverec2([ca1, (ch1, cv1, cd1)], 'db2') waterImg = np.array(waterImg, np.uint8) waterImg = self.deArnold(waterImg) kernel = np.ones((3, 3), np.uint8) if flag == 0: waterImg = cv2.erode(waterImg, kernel) elif flag == 1: waterImg = cv2.dilate(waterImg, kernel) cv2.imwrite('水印.png', waterImg) return waterImg if __name__ == '__main__': img = 'a.png' k = 20 xs = [0.2, 0.2, 0.5, 0.4] W1 = WaterMarkDWT(img, waterImg, k, xs)
3):编写exp.py得到flag.png:
import cv2 import pywt import numpy as np class WaterMarkDWT: def __init__(self, origin: str, watermark: str, key: int, weight: list): self.key = key self.img = cv2.imread(origin) self.mark = cv2.imread(watermark) self.coef = weight def arnold(self, img): r, c = img.shape p = np.zeros((r, c), np.uint8) a, b = 1, 1 for k in range(self.key): for i in range(r): for j in range(c): x = (i + b * j) % r y = (a * i + (a * b + 1) * j) % c p[x, y] = img[i, j] return p def deArnold(self, img): r, c = img.shape p = np.zeros((r, c), np.uint8) a, b = 1, 1 for k in range(self.key): for i in range(r): for j in range(c): x = ((a * b + 1) * i - b * j) % r y = (-a * i + j) % c p[x, y] = img[i, j] return p def get(self, size: tuple = (1200, 1200), flag: int = None): img = cv2.resize(self.img, size) img1 = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY) img2 = cv2.cvtColor(self.mark, cv2.COLOR_RGB2GRAY) c = pywt.wavedec2(img2, 'db2', level=3) [cl, (cH3, cV3, cD3), (cH2, cV2, cD2), (cH1, cV1, cD1)] = c d = pywt.wavedec2(img1, 'db2', level=3) [dl, (dH3, dV3, dD3), (dH2, dV2, dD2), (dH1, dV1, dD1)] = d a1, a2, a3, a4 = self.coef ca1 = (cl - dl) * a1 ch1 = (cH3 - dH3) * a2 cv1 = (cV3 - dV3) * a3 cd1 = (cD3 - dD3) * a4 waterImg = pywt.waverec2([ca1, (ch1, cv1, cd1)], 'db2') waterImg = np.array(waterImg, np.uint8) waterImg = self.deArnold(waterImg) kernel = np.ones((3, 3), np.uint8) if flag == 0: waterImg = cv2.erode(waterImg, kernel) elif flag == 1: waterImg = cv2.dilate(waterImg, kernel) cv2.imwrite('水印.png', waterImg) return waterImg if __name__ == '__main__': img = 'a.png' newImg='newImg.png' k = 20 #xs = [0.2, 0.2, 0.5, 0.4] coef=[5,5,2,2.5] #waterImg='flag.png' W1 = WaterMarkDWT(img, newImg, k, coef) waterimg=W1.get()
第六题:Reverse逆向RC4
6.1 题目文件rc4.exe如下所示:
6.2 解题步骤:
1):die检测无壳,32位程序。
2):用IDA32打开,发现key为gamelab@,提取数组v5进行RC4解密得到flag。
第七题:ECDSA椭圆曲线数字签名算法
7.1 题目代码如下所示:
import ecdsa import random def ecdsa_test(dA,k): sk = ecdsa.SigningKey.from_secret_exponent( secexp=dA, curve=ecdsa.SECP256k1 ) sig1 = sk.sign(data=b'Hi.', k=k).hex() sig2 = sk.sign(data=b'hello.', k=k).hex() r1 = int(sig1[:64], 16) s1 = int(sig1[64:], 16) s2 = int(sig2[64:], 16) return r1,s1,s2 if __name__ == '__main__': n = 0xfffffffffffffffffffffffffffffffebaaedce6af48a03bbfd25e8cd0364141 a = random.randint(0,n) flag = 'flag{' + str(a) + "}" b = random.randint(0,n) print(ecdsa_test(a,b)) # (4690192503304946823926998585663150874421527890534303129755098666293734606680, 111157363347893999914897601390136910031659525525419989250638426589503279490788, 74486305819584508240056247318325239805160339288252987178597122489325719901254)
7.2 代码分析:
椭圆曲线数字签名算法,它利用椭圆曲线密码学(ECC)对数字签名算法(DSA)进行模拟,其安全性基于椭圆曲线离散对数问题。但是当某些数值相同时会出现一些安全问题。
分析代码可以看出,存在随机数重复使用。具体来说,这段代码中签名的过程中使用了相同的随机数 k 来对不同的消息进行签名。这种情况下,可以通过分析两个相同 k 值对应的消息签名来恢复私钥 dA。
在 ECDSA 中,每次签名过程中都会使用一个随机数 k,以确保生成唯一的签名。然而,如果相同的随机数 k 被重复使用来对不同的消息进行签名,攻击者就有可能通过数学分析和推导计算出私钥 dA。
7.3 exp代码:
import sympy from hashlib import sha1 from Cryptodome.Util.number import long_to_bytes , bytes_to_long def calculate_private_key(r1, s1, s2, h1, h2, n): # 计算k值 k = ((h1 - h2) * sympy.mod_inverse(s1 - s2, n)) % n # 计算私钥dA dA = (sympy.mod_inverse(r1, n) * (k * s1 - h1)) % n return dA if __name__ == "__main__": # 定义椭圆曲线的参数 n = 0xfffffffffffffffffffffffffffffffebaaedce6af48a03bbfd25e8cd0364141 # 签名中的r1, s1, s2值 r1 = 4690192503304946823926998585663150874421527890534303129755098666293734606680 s1 = 111157363347893999914897601390136910031659525525419989250638426589503279490788 s2 = 74486305819584508240056247318325239805160339288252987178597122489325719901254 h1 = bytes_to_long(sha1(b'Hi.').digest()) h2 = bytes_to_long(sha1(b'hello.').digest()) private_key = calculate_private_key(r1, s1, s2, h1, h2, n) print(f'flag{{{private_key}}}')
获得flag{40355055231406097504270940121798355439363616832290875140843417522164091270174}
第八题:Reverse逆向xxtea
8.1 题目描述如下:
Flag被使用了算法分成若干个小块,每个块使用相同的加密解密方法,但这个算法是对称加密,请分析密文并还原。
8.2 题目文件happytime
8.3 解题步骤
1):die检测无壳,64位程序。
2):拖进IDA分析。
printf输出提示信息Let’s have a drink,pay your answer(flag):,read在键盘读取flag输入,重要的关键函数是cry加密函数,接受v5和输入的flag,这里的11应该是flag被分割成了11组,最后一个循环比较加密后的flag和V6,刚好和上面v6数组对应:
3):根据cry函数的特征,可以判定这是XXTEA加密无疑,找到其中的DELTA,密文(main函数中的v6),和key(前面main函数的V5),即可编写脚本解密(输出的时候注意大小端序):
#include #include #define MX (((z >> 5) ^ (y > 3) ^ (z 1) { /* enCoding Part */ q = 415 / n + 114; while (q-- > 0) { sum += DELTA; e = (sum >> 2) & 3; for (p = 0; p > 2) & 3; for (p = n - 1; p > 0; p--) { z = v[p - 1]; y = v[p] -= MX; } z = v[n - 1]; y = v[0] -= MX; sum += DELTA; } return 0; } return 1; } int main() { unsigned int v[11] = {0x480AC20C, 0xCE9037F2, 0x8C212018, 0xE92A18D, 0xA4035274, 0x2473AAB1, 0xA9EFDB58, 0xA52CC5C8, 0xE432CB51, 0xD04E9223, 0x6FD07093}, key[4] = {0x79696755, 0x67346F6C, 0x69231231, 0x5F674231}; int n = 11; // n为要加密的数据个数 btea(v, -n, key); // 取正为加密,取负为解密 char *p = (char *)v; for (int i = 0; i恭喜CTF培训班的王思雨同学获得2024年蓝桥杯省赛一等奖,国赛三等奖的成绩。