<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>Posts on Akshit Kathuria</title>
    <link>https://ak-debug.github.io/posts/</link>
    <description>Recent content in Posts on Akshit Kathuria</description>
    <image>
      <title>Akshit Kathuria</title>
      <url>https://ak-debug.github.io/social-preview.png</url>
      <link>https://ak-debug.github.io/social-preview.png</link>
    </image>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Thu, 12 Nov 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://ak-debug.github.io/posts/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Broken Access Control: IDOR and Cookie Tampering</title>
      <link>https://ak-debug.github.io/posts/broken-access-control-idor-and-cookies/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/broken-access-control-idor-and-cookies/</guid>
      <description>When the server trusts a value the client can change to decide who you are or what you can see, you can often just change it. Here I cover two sides of broken access control: tampering with a client-side cookie to become an admin, and incrementing an ID in the URL to read someone else&amp;rsquo;s record, using two DD CTF challenges as worked examples.</description>
    </item>
    <item>
      <title>Bypassing CSRF Token Randomization Using an XSS Flaw</title>
      <link>https://ak-debug.github.io/posts/bypassing-csrf-token-randomization-with-xss/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/bypassing-csrf-token-randomization-with-xss/</guid>
      <description>Anti-CSRF tokens stop a classic cross-site request forgery attack, but they assume the attacker can&amp;rsquo;t read the page. An XSS vulnerability breaks that assumption. Here I chain a stored XSS flaw on the SEED Labs Elgg platform to exfiltrate a victim&amp;rsquo;s session cookie and CSRF tokens, then replay any authenticated action as them.</description>
    </item>
    <item>
      <title>Common RSA Weaknesses: Shared Primes, Close Primes, and Bad Parameters</title>
      <link>https://ak-debug.github.io/posts/rsa-parameter-weaknesses/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/rsa-parameter-weaknesses/</guid>
      <description>RSA is only as strong as the primes behind it. If two keys share a prime, or if p and q sit too close together, the modulus can be factored in seconds with a GCD or Fermat&amp;rsquo;s method. Here I explain the maths behind each weakness, with clean Python solves for four crypto challenges from the bi0s DD CTF.</description>
    </item>
    <item>
      <title>Detecting Code Injection in Memory Dumps: SilentBanker and Coreflood</title>
      <link>https://ak-debug.github.io/posts/detecting-code-injection-in-memory-dumps/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/detecting-code-injection-in-memory-dumps/</guid>
      <description>Banking trojans rarely run as their own process. They copy themselves into a browser and hook its functions, so the process list looks clean. Here I walk through a Volatility workflow for finding injected code in a Windows memory image: cross-checking process views, YARA hunting, malfind, apihooks, impscan, and connscan. The worked examples are Coreflood and SilentBanker, both injected into Internet Explorer.</description>
    </item>
    <item>
      <title>Enumerating a Server Through an Exposed Command Interface</title>
      <link>https://ak-debug.github.io/posts/enumerating-a-server-through-command-execution/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/enumerating-a-server-through-command-execution/</guid>
      <description>When a web app hands you the ability to run shell commands, the challenge isn&amp;rsquo;t getting code execution, it&amp;rsquo;s knowing how to look around. Here I cover methodical filesystem enumeration from a command primitive, including seeing past a decoy flag to a hidden one, and why exposing command execution is total compromise.</description>
    </item>
    <item>
      <title>Exploiting XXE: Reading Local Files via XML External Entities</title>
      <link>https://ak-debug.github.io/posts/exploiting-xxe-to-read-local-files/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/exploiting-xxe-to-read-local-files/</guid>
      <description>When an application parses XML you control and its parser resolves external entities, you can make the server read local files and hand them back to you. This is XXE. Here I explain why XML parsers are dangerous by default, walk through reading a file off the server with a classic XXE payload, and cover how to shut it down.</description>
    </item>
    <item>
      <title>From File Upload to Remote Code Execution</title>
      <link>https://ak-debug.github.io/posts/file-upload-to-rce/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/file-upload-to-rce/</guid>
      <description>If a web server lets you upload a file and will later execute it, an upload form becomes a way to run your own code on the server. Here I walk through confirming PHP execution with a harmless test payload and then reading a protected file, plus why &amp;lsquo;store the upload in the webroot&amp;rsquo; is the real mistake.</description>
    </item>
    <item>
      <title>Image Forensics: Metadata, Hidden Dimensions, and Layer Analysis</title>
      <link>https://ak-debug.github.io/posts/image-forensics-metadata-and-hidden-dimensions/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/image-forensics-metadata-and-hidden-dimensions/</guid>
      <description>An image can hide data in its header, in its metadata, or in the pixel values themselves. Here I cover three independent techniques for finding it: rewriting the dimensions in a JPEG&amp;rsquo;s SOF0 marker to reveal cropped-out rows, reading metadata with exiftool and decoding base64, and splitting an image into bit planes with Aperi&amp;rsquo;Solve. Three forensics challenges from the bi0s DD CTF serve as worked examples.</description>
    </item>
    <item>
      <title>Memory Forensics with Volatility: Reading What a RAM Dump Remembers</title>
      <link>https://ak-debug.github.io/posts/memory-forensics-with-volatility/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/memory-forensics-with-volatility/</guid>
      <description>A RAM dump captures what a Windows machine was doing at one moment: running processes, open files, console history, browser history, and environment variables. This guide covers a repeatable Volatility 2 workflow, from identifying the profile with imageinfo through process listing, file recovery, console and browser history, and dumping a process for strings. Three memory forensics challenges from the bi0s DD CTF serve as worked examples.</description>
    </item>
    <item>
      <title>Path Traversal and LFI: Reading Files the App Never Meant to Serve</title>
      <link>https://ak-debug.github.io/posts/local-file-inclusion-and-path-traversal/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/local-file-inclusion-and-path-traversal/</guid>
      <description>When a web app takes a filename from the user and reads it off disk, a few ../ sequences can walk out of the intended directory and into anything the server can read. Here I cover path traversal end to end, including a filter-bypass trick for when the app tries to strip dangerous input, using two DD CTF challenges as worked examples.</description>
    </item>
    <item>
      <title>Reconstructing Data from Packet Captures</title>
      <link>https://ak-debug.github.io/posts/reconstructing-data-from-packet-captures/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/reconstructing-data-from-packet-captures/</guid>
      <description>A file sent over the network never shows up as one neat packet. It is split across many packets, and to recover it you have to pick out the right ones, put them in order, and join their payloads. Here I cover how that works for TCP streams and for data smuggled one byte at a time through ICMP echo payloads, using Wireshark and Scapy on four network-forensics challenges from the bi0s DD CTF.</description>
    </item>
    <item>
      <title>Repairing Corrupted PNG Files by Hand</title>
      <link>https://ak-debug.github.io/posts/repairing-corrupted-png-files/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/repairing-corrupted-png-files/</guid>
      <description>A PNG that refuses to open is rarely beyond repair. Its structure is simple: an 8-byte signature followed by length-type-data-CRC chunks. Once you know that layout, you can fix a broken file in a hex editor. Here I walk through repairing a mangled signature, IHDR header, chunk name, and CRC checksums, using two forensics challenges from the bi0s DD CTF as worked examples.</description>
    </item>
    <item>
      <title>Solving Classical and Homebrew Ciphers</title>
      <link>https://ak-debug.github.io/posts/solving-classical-and-homebrew-ciphers/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/solving-classical-and-homebrew-ciphers/</guid>
      <description>Most CTF cipher challenges come down to a few ideas: modular arithmetic, substitution with a repeating key, reused XOR keystreams, and historical machines. This post covers how I identify which one I&amp;rsquo;m facing, then works through four challenges from the bi0s DD CTF: a modulo &amp;lsquo;hash&amp;rsquo; that can be inverted, an Enigma message, a broken AES-CTR implementation, and a Vigenère cipher with an extended alphabet.</description>
    </item>
    <item>
      <title>SQL Injection: Bypassing a Login with OR 1=1</title>
      <link>https://ak-debug.github.io/posts/sql-injection-login-bypass/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/sql-injection-login-bypass/</guid>
      <description>A login form that builds its SQL query by pasting your input into a string can be talked into authenticating you as someone else. Here I walk through turning a SQL error into a login bypass with a classic OR &amp;lsquo;1&amp;rsquo;=&amp;lsquo;1&amp;rsquo; payload, why it works at the query level, and how parameterised queries end the whole bug class.</description>
    </item>
    <item>
      <title>Stealing an Admin Session with XSS and an Out-of-Band Listener</title>
      <link>https://ak-debug.github.io/posts/stealing-admin-session-with-stored-xss/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/stealing-admin-session-with-stored-xss/</guid>
      <description>A feedback form that reflects your input and an admin who visits the links you submit is everything you need to steal the admin&amp;rsquo;s session cookie. Here I chain an XSS payload with an out-of-band listener so the victim&amp;rsquo;s browser sends their cookie to me, and explain the admin-bot pattern and the defences that break it.</description>
    </item>
    <item>
      <title>Steganography: Finding Data Hidden in Images and Whitespace</title>
      <link>https://ak-debug.github.io/posts/steganography-extracting-hidden-data/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/steganography-extracting-hidden-data/</guid>
      <description>Steganography hides data in plain sight, inside an ordinary image or an innocent-looking text file. This is a reference guide to the standard workflow: inspect a file with aperisolve, exiftool, binwalk, and strings; pull embedded data out of an image with steghide; brute-force an unknown passphrase with a wordlist; and detect and decode whitespace stego with stegsnow. Two forensics challenges from the bi0s DD CTF serve as worked examples.</description>
    </item>
    <item>
      <title>Web Recon: Finding Secrets in Source, robots.txt, and Client-Side Storage</title>
      <link>https://ak-debug.github.io/posts/web-recon-finding-client-side-secrets/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/web-recon-finding-client-side-secrets/</guid>
      <description>Developers keep hiding data in places the browser can read: HTML comments, robots.txt, hidden fields, and local storage. None of it is secret. Here I reassemble a flag split across the page source, a robots.txt disallow entry, and an inspected value, and explain why client-side is never a hiding place.</description>
    </item>
    <item>
      <title>When Symmetric Crypto Goes Wrong: AES Key-as-IV and XOR Mistakes</title>
      <link>https://ak-debug.github.io/posts/symmetric-crypto-misuse-aes-and-xor/</link>
      <pubDate>Thu, 12 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://ak-debug.github.io/posts/symmetric-crypto-misuse-aes-and-xor/</guid>
      <description>AES is not broken, but code that uses it often is. Here I explain why reusing the key as the CBC initialisation vector can hand an attacker the key itself, and why home-made XOR ciphers fall to known plaintext and key reuse. Each mistake comes with a working Python demonstration, the correct pattern, and a worked example from the bi0s DD CTF.</description>
    </item>
  </channel>
</rss>
