<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by MahmoudKroush (xgoon) on Medium]]></title>
        <description><![CDATA[Stories by MahmoudKroush (xgoon) on Medium]]></description>
        <link>https://medium.com/@kroush333?source=rss-e61e334e1ece------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*LciwzaF0RH_njeQ7</url>
            <title>Stories by MahmoudKroush (xgoon) on Medium</title>
            <link>https://medium.com/@kroush333?source=rss-e61e334e1ece------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Mon, 16 Nov 2026 23:39:45 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@kroush333/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Stack Your Arsenal: Logic Bugs That Turn Into $$$]]></title>
            <link>https://medium.com/@kroush333/stack-your-arsenal-logic-bugs-that-turn-into-76ef49973b70?source=rss-e61e334e1ece------2</link>
            <guid isPermaLink="false">https://medium.com/p/76ef49973b70</guid>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[logic-bug]]></category>
            <category><![CDATA[bug-bounty]]></category>
            <category><![CDATA[business-logic-bug]]></category>
            <category><![CDATA[bug-hunting]]></category>
            <dc:creator><![CDATA[MahmoudKroush (xgoon)]]></dc:creator>
            <pubDate>Sun, 15 Mar 2026 14:06:35 GMT</pubDate>
            <atom:updated>2026-03-15T14:06:35.290Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SRBMlosp3SrV8UgvMVMM_g.png" /></figure><p>بسم الله، والصلاة والسلام على رسول الله<br>الحمد لله الذي عَلَّمَ بالقلم.. عَلَّمَ الإنسانَ ما لم يَعْلَم والصلاةُ والسلامُ على خيرِ مُعَلِّمي الناسِ الخير محمد</p><h3>Who am I</h3><p>I’m Mahmoud Kroush (xgoon), a recently graduated from the Faculty of Engineering. I’m aiming to become a penetration tester and currently work part-time as a bug bounty hunter.</p><blockquote>What are business logic vulnerabilities?</blockquote><blockquote>Business logic vulnerabilities are flaws in the design and implementation of an application that allow an attacker to elicit unintended behavior. This potentially enables attackers to manipulate legitimate functionality to achieve a malicious goal. These flaws are generally the result of failing to anticipate unusual application states that may occur and, consequently, failing to handle them safely.</blockquote><p>In brief, logic bugs usually mean bypassing restrictions. When an application tells you <em>‘you can’t do this’</em>, it can be a hint that a logic bug exists — something that allows you to turn a <em>No</em> into a <em>Yes</em>.</p><p>In this write-up, I’ll dive into some of the logic bugs I found during the past few months of hunting. These examples are simple, but they can be useful for new hunters to understand what to look for when searching for logic bugs.</p><h3>1. Bypass Required Fields</h3><p>While working on an external application that offered a CRM, I found a logic bug in the “Leads” section. Each lead had user-related data, including an <strong>Email</strong> field marked as <em>required</em>. Normally, if you leave it empty, the app prevents you from saving.</p><p>So I thought: <strong>If the app says “you can’t,” let’s try to bypass it.</strong></p><p>First, I tried injecting an HTML tag like &lt;script&gt; while creating a lead. The app correctly encoded the characters (&lt; → &amp;lt;), so no issue there.<br> But when I saved the lead and then <strong>edited</strong> it, I noticed something interesting:</p><p>When I inserted &lt;script&gt; into the field during editing, the application <strong>stripped the entire tag</strong> and left the field <strong>empty</strong>—and still allowed me to save it.That means I could bypass the “required field” validation completely.I wrote the report, submitted it</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zNhj90nYHwCcph2XMIwMRw.jpeg" /></figure><h3>2. Using a “Premium” Feature Without Permission</h3><p>In the CRM, the app had a feature called <strong>Contacts</strong>, where users could store contact information. I created two accounts:</p><ul><li>one <strong>trial</strong> account</li><li>one <strong>free</strong> account (which normally has <em>no</em> access to the CRM)</li></ul><p>The trial user invited the free user into his CRM workspace, and I noticed something interesting:</p><p>The <strong>Contacts</strong> feature suddenly became available for the free user.</p><p>Then, using the trial account, I deleted the free user from the CRM. After that, the free user went back to the normal state — <strong>no CRM access</strong>.</p><p>But here’s the bug:</p><p>When I checked Burp, I found that the free user could still send requests to <strong>create</strong>, <strong>edit</strong>, and <strong>export</strong> Contacts.<br> Even though the UI blocked access, the backend still allowed these premium actions.</p><p>So the free account was able to use a <strong>premium feature</strong> just by replaying the API requests.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*T9b98Fpv39WQDw38ZM-tzg.jpeg" /></figure><h3>3. Bypass the Maximum Number of Pages</h3><p>You might think this one is a race condition — but <strong>NO</strong> :).</p><p>The app lets users create websites with a limited number of pages.<br> For free users, the limit is 5<strong>0 pages</strong>.<br> I tried the usual race-condition trick, but it didn’t work (looks like they already fixed that).</p><p>But I found a much simpler flow to bypass the limit:</p><p>A user can delete pages, and the deleted pages move to the <strong>trash</strong>. From there, you can <strong>restore</strong> them.</p><p>Here’s the bug:<br> When restoring a deleted page, the app <strong>does not check</strong> how many active pages the user currently has.</p><p>So I created 50 pages (the max), then went to the Trash and started restoring the deleted ones.<br> They restored <strong>normally</strong>, and I ended up with <strong>more than 50 live pages</strong>.</p><p>A clean logic bug from a missing backend validation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/619/1*BBCDErhRDiSkAlgMI0PZNw.png" /></figure><h3>4. Editing a Deleted Product</h3><p>In the same application, there was a feature to create a store and add products. Normally, when you move a product to the <strong>trash</strong>, you <strong>shouldn’t</strong> be able to edit it anymore. You must restore it first before editing.</p><p>But “you can’t” → <strong>logic bug time</strong>.</p><p>So here’s what I did:</p><p>I created a new product, edited it normally, and captured the request in Burp.<br> Then I deleted the product (moved it to trash).<br> After that, I went back to Burp and replayed the <strong>edit</strong> request from Repeater Find <strong>still accepted the edit</strong> request.</p><p>So I was able to <strong>edit a deleted product</strong> without restoring it — a clear logic flaw.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/625/1*nEvlKIAeH5Kgg_ujx0hj0Q.png" /></figure><h3>5. Move Pages to a Deleted Folder</h3><p>In the site builder, users can create <strong>Folders</strong> and move pages inside them. When you delete a folder, you <strong>shouldn’t</strong> be able to access or edit anything inside it anymore.</p><p>But “you can’t” again? <strong>My time.</strong></p><p>I noticed that each folder had two key actions in the API:</p><ol><li><strong>Edit folder</strong></li><li><strong>Move page to folder</strong></li></ol><p>After deleting a folder, the first action (edit) correctly returned an error — no access.</p><p>But the second one?<br> When I tried <strong>“move page to folder”</strong> on a deleted folder, the server responded with <strong>200 OK</strong> and successfully performed the action.</p><p>So even though the folder was in the trash and not accessible through the UI, the backend still allowed pages to be moved into it.</p><p>This means a user could <strong>hide pages inside a deleted folder</strong>, making them unreachable from the interface — another nice logic flaw caused by missing state validation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/607/1*Z2MUWjCYtqroO635eNxoHA.png" /></figure><h3>6. Render Failure Leading to Crashing the Entire Presentation</h3><p>While I was working on a private program on HackerOne, I found a logic bug in an application responsible for creating and presenting slide presentations.</p><p>Inside the editor, you can change the presentation’s bakcground_color theme. The request for editing the color looked like this:</p><pre>{<br>  &quot;background_color&quot;: &quot;rgb(102,50,70)&quot;<br>}</pre><p>The <strong>RGB</strong> value represents the Red, Green, and Blue components.<br> The server takes this RGB value, performs some calculations, and generates the final color.</p><p>While testing, I changed the color value to random inputs and noticed that the server <strong>accepted</strong> whatever I sent.<br> But when I returned to the UI and tried to reload the presentation, the entire presentation <strong>crashed</strong> — nobody could access it anymore.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*33MYqBAHUrNHsGdfmzDM4w.png" /></figure><p><strong>Why did this happen?</strong></p><p>The developer relied completely on the value coming from the client-side.<br> They passed the RGB string directly into the color calculation function <strong>without validating</strong> whether the value was valid or even properly formatted.</p><p>A simple invalid color value sent by the client caused the render function to fail, which resulted in the entire presentation breaking for all users — a clear logic flaw coming from trusting user input too much.</p><h3>7. Tampering With Logs</h3><p>While working on a private program on Bugcrowd, I found a logic bug related to how the target handled logging.<br> This application relied heavily on logs to track every action done by each user.</p><p>When I tested the feature for creating a new field, I noticed that the database stored a value called <strong>created_by</strong>.<br> So I intercepted the request and changed the <strong>created_by</strong> email to any user I wanted.</p><p>The result?<br> It <strong>accepted</strong> the request, and the system successfully created the action under another user’s email.</p><p>Why did this happen?</p><p>The developer trusted the <strong>created_by</strong> parameter coming from the client-side request body.<br> Instead of using the authenticated <strong>session token</strong> to determine who performed the action, the backend relied entirely on user-supplied data.</p><p>This allowed me to tamper with the logs and make it look like <strong>any user</strong> performed the action — a clear logic flaw caused by trusting client-controlled parameters.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/817/1*TkNfpUDJdWdaju5Z474NJA.png" /></figure><h3>8. Editing Preset Roles Inside an Organization</h3><p>When creating a new organization in the application, you’ll find some <strong>preset roles</strong> already available — like <em>owner</em>, <em>reader</em>, etc.<br> Users can also create <strong>custom roles</strong>, and these can be edited freely.</p><p>But when you try to edit a <strong>preset role</strong>, the UI disables the edit button and shows a client‑side alert saying that this role cannot be edited.</p><p>So I created a new custom role, opened the edit page, and intercepted the request.<br> Then I simply <strong>replaced the custom role ID with the preset role ID</strong>.</p><p>The result?<br> The server <strong>accepted</strong> the request, and the preset role was successfully edited.</p><p>The logic flaw here is obvious:<br> The restriction was only enforced on the <strong>client side</strong>, with no backend validation to block editing preset roles.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/820/1*uDLKLZGheMiryd7m6vEXlg.png" /></figure><p>🤍في الاخر اتمني تكون استفدت باي حاجه و متنساش تدعي ل اخواتنا المسلمين في كل بقاع الارض</p><p><a href="http://www.linkedin.com/in/mahmoud-kroush-2a4011289">www.linkedin.com/in/mahmoud-kroush-2a4011289</a><br><a href="https://x.com/MahmoudKroush33">https://x.com/MahmoudKroush33</a><br>“سبحانك اللهم وبحمدك، أشهد أن لا إله إلا أنت، أستغفرك وأتوب إليك”.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=76ef49973b70" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Journey from FOFA Dorking to Critical Remote Access]]></title>
            <link>https://medium.com/@kroush333/journey-from-fofa-dorking-to-critical-remote-access-b337f92f3d28?source=rss-e61e334e1ece------2</link>
            <guid isPermaLink="false">https://medium.com/p/b337f92f3d28</guid>
            <category><![CDATA[bug-bounty]]></category>
            <category><![CDATA[reconnaissance]]></category>
            <category><![CDATA[bugbounty-writeup]]></category>
            <dc:creator><![CDATA[MahmoudKroush (xgoon)]]></dc:creator>
            <pubDate>Fri, 05 Dec 2025 15:54:21 GMT</pubDate>
            <atom:updated>2026-03-30T11:29:32.492Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/900/1*zz4vhgt80G1QwccCLdqysw.png" /></figure><p>بسم الله، والصلاة والسلام على رسول الله<br>الحمد لله الذي عَلَّمَ بالقلم.. عَلَّمَ الإنسانَ ما لم يَعْلَم والصلاةُ والسلامُ على خيرِ مُعَلِّمي الناسِ الخير محمد</p><h3>Who am I</h3><p>I’m Mahmoud Kroush, a recently graduated student from the Faculty of Engineering. I’m aiming to become a penetration tester and currently work part-time as a bug bounty hunter.</p><h3><strong>Introduction</strong></h3><p>At the beginning of the month, I was experiencing burnout and cloudn’t focus on new targets. I decided to improve my skills in reconnaissance — an area I hadn’t particularly enjoyed before. I spent time reading various write-ups, gathering information and tips that I’ll share in this article. This learning journey ultimately led me to discover a critical bug.</p><h3><strong>First Step:</strong></h3><p>I wanted to activate my December streak on HackerOne. I had a targets in mind that I regularly check during reconnaissance tests, where bugs often remain unresolved — let’s call it Redacted.com. I went to the Fofa search engine and used this dork:</p><pre>domain=&quot;Redacted.com&quot; &amp;&amp; body=&quot;secret&quot;</pre><p>This search queries all live hosts under the target domain for the word “secret” in the body content. I also tested other critical keywords like “config” and similar terms Look into this writeups.</p><blockquote><a href="https://medium.com/legionhunters/fofa-dorking-for-bug-hunters-a35c80bbab6e">https://medium.com/legionhunters/fofa-dorking-for-bug-hunters-a35c80bbab6e</a></blockquote><blockquote><a href="https://medium.com/@GERRR4Y/recon-like-a-hunter-practical-tips-from-real-findings-part-2-e71bcf5af0ae">https://medium.com/@GERRR4Y/recon-like-a-hunter-practical-tips-from-real-findings-part-2-e71bcf5af0ae</a></blockquote><p>While this process could be boring, it sometimes yields very important findings.</p><p>I discovered a subdomain — let’s say vnc-infra-redacted.redacted.com. When I viewed the page source and searched for the word “secret,” I found it embedded in JavaScript code.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KoaIa0SfmqwmwPRN0Dbnkg.png" /></figure><h3><strong>Second Step:</strong></h3><p>While analyzing the code, I found two variables needed to complete the path: namespace and reservation. I could concatenate these to construct the full URL. However, something caught my attention—both the subdomain and the path contained the word &quot;VNC.&quot;</p><p><strong>What is VNC?</strong></p><p>When I encounter something unfamiliar, I follow a three-step process:</p><ol><li><strong>Ask AI</strong> (ChatGPT, Claude) to give me information about the service</li><li><strong>Search for disclosed reports</strong> on HackerOne related to this service</li><li><strong>Google it</strong> to gather more information from different resources</li></ol><p><strong>What is VNC?</strong></p><blockquote>VNC (Virtual Network Computing) is a remote desktop protocol that allows a user to view and control another computer’s screen over the network. It’s commonly used to manage servers and virtual machines. If exposed without proper authentication, it can allow attackers to see and interact with the machine’s desktop, leading to full system compromise.</blockquote><h3><strong>Third Step:</strong></h3><p>I copied the first endpoint from the URL: /ksdq/api/v1/namespaces/, which gave me access to data about running namespaces.I extracted several namespace IDs from the response.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/820/1*7yi25keZ9UAfdhhW2_K_FQ.png" /></figure><p>Searching further through the JavaScript code, I found similar patterns and discovered that the reservation variable takes the same value as the namespace ID. Now I could construct the full path:</p><pre>https://vnc-infra-redacted.redacted.com/ksdq/api/v1/namespaces/namespace_id/secrets/namespaceid-vnc</pre><p>This redirected me to a JSON page containing information about the namespace, including a secret token.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/597/1*UFIN0sXeAZzh-zYVyEMUGw.png" /></figure><p>I had found interesting data, but I wanted to go deeper — specifically, to gain remote desktop access. However, the current information wasn’t sufficient. I went back to the JavaScript code and discovered another endpoint.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*po2VBIJO9FYPcCI3bP2pNg.png" /></figure><p>I needed two variable: the apiPrefix = (ksdq/apis) and the namespace ID. Combining these, the full URL became:</p><pre>https://vnc-infra-redacted.redacted.com/ksdq/apis/redacted.io/v1/namespaces/namespace_id/virtualmachineinstances</pre><p>This led me to a JSON page containing additional data, from which I extracted a new secret key.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/508/1*lvkTeqGxAiLrKtBBqpUbGQ.png" /></figure><h3><strong>Fourth Step:</strong></h3><p>I collected all the data and continued researching VNC technology to understand how to gain access. I discovered that the final link would be:</p><pre>https://vnc-infra-redacted.redacted.com/vnc.html?autoconnect=true&amp;path=ksdq/apis/subresources.redacted.io/v1/namespaces/namespace_id/virtualmachineinstances/secret_key/vnc</pre><p>This brought me to a virtual machine access interface where I could choose to log in as either a user or an administrator.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*X_BD1HnXmHORR3RzjpqBLg.png" /></figure><h3><strong>Finally:</strong></h3><p>I compiled all my findings and submitted a detailed report to the security team. After their review, they triaged the vulnerability as <strong>Critical</strong> due to the potential for unauthorized remote desktop access to virtual machine instances.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/432/1*4dpkpgalUscA55pP8zZIaw.png" /></figure><p>🤍في الاخر اتمني تكون استفدت باي حاجه و متنساش تدعي ل اخواتنا المسلمين في كل بقاع الارض</p><p><a href="http://www.linkedin.com/in/mahmoud-kroush-2a4011289">www.linkedin.com/in/mahmoud-kroush-2a4011289</a><br><a href="https://x.com/MahmoudKroush33">https://x.com/MahmoudKroush33</a><br>“سبحانك اللهم وبحمدك، أشهد أن لا إله إلا أنت، أستغفرك وأتوب إليك”.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b337f92f3d28" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Be careful — I could steal your identity | Business logic bug]]></title>
            <link>https://medium.com/@kroush333/be-careful-i-could-steal-your-identity-business-logic-bug-76c1d83868dc?source=rss-e61e334e1ece------2</link>
            <guid isPermaLink="false">https://medium.com/p/76c1d83868dc</guid>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[hackerone]]></category>
            <category><![CDATA[business-logic]]></category>
            <category><![CDATA[bug-bounty]]></category>
            <category><![CDATA[bugbounty-writeup]]></category>
            <dc:creator><![CDATA[MahmoudKroush (xgoon)]]></dc:creator>
            <pubDate>Wed, 12 Nov 2025 11:53:58 GMT</pubDate>
            <atom:updated>2025-11-13T08:11:02.455Z</atom:updated>
            <content:encoded><![CDATA[<h3>Be careful — I could steal your identity | Business logic bug</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fDGeHlluyINFjMq8OYrwQA.jpeg" /></figure><p>بسم الله، والصلاة والسلام على رسول الله <br>الحمد لله الذي عَلَّمَ بالقلم.. عَلَّمَ الإنسانَ ما لم يَعْلَم والصلاةُ والسلامُ على خيرِ مُعَلِّمي الناسِ الخير محمد</p><h3>Who am I</h3><p>I’m Mahmoud Kroush, a recently graduated student from the Faculty of Engineering. I’m aiming to become a penetration tester and currently work part-time as a bug bounty hunter.</p><h3>Introduction:</h3><p>This write-up examines a misconfiguration in the “change email” feature that could allow an attacker to hijack any user’s identity on the platform. The program is a private HackerOne disclosure; for privacy I’ll refer to it as Redicted.com.</p><h3>Brief About the Target:</h3><p>The target is a web platform that helps users create and present professional presentations. Platform owners can create teams for collaborative editing and presentation. I’ll skip lengthy background , Let’s go to Bug</p><h3>Frist Step:</h3><p>After exploring the target and learning how it works, I started testing authentication. When I reached the change-email feature I found the process goes like this:</p><ol><li>Set new email, confirm new email, enter current password.</li><li>The site redirects you to /user/verify to enter the verification token sent to the new email.</li><li>Enter the OTP that was sent to the email inbox and the change of email is confirmed.</li></ol><h3>Second Step:</h3><p>After I entered the new email (new_email_one@gmail.com) and my password, I intercepted the request, sent it to Repeater, and forwarded it. I confirmed the verification code was sent to new_email_one@gmail.com. I then tried a simple test: I changed the email parameter to an array of emails ([&quot;new_user_one@gmail.com&quot;, &quot;new_email_two@gmail.com&quot;]) and sent the request.</p><pre>PUT /users HTTP/2<br>Host: api.redicted.com<br>Content-Type: application/json<br>Authorization: Bearer &lt;REDACTED&gt;<br><br>{ &quot;email&quot;: [new_user_one@gmail.com , new_email_two@gmail.com], &quot;flags&quot;: {}, &quot;password&quot;: &quot;*****&quot; }</pre><p>i find that server accept the email and Response with 200 OK</p><pre>HTTP/2 200 OK<br><br>{ &quot;id&quot;:xxxx, &quot;user_id&quot;:xxxx, &quot;created_at&quot;:&quot;2025-09-30T13:07:33+00:00&quot;, &quot;name&quot;:&quot;user_burble&quot;, &quot;email&quot;:&quot;[\&quot;new_user_one@gmail.com\&quot;, \&quot;new_email_two@gmail.com\&quot;]&quot;, ... , &quot;is_verified&quot;:false, &quot;requires_email_verification&quot;:true }</pre><p>And the verification token is sent to (new_email_one@gmail.com). OK, first I thought that the email was sent to the first email in the array only, but when I changed the order of the array it also sent the verification token to (new_email_one@gmail.com) wait a second. What is happening here?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/500/1*sw4PkONUssJQt3_3SxAIug.gif" /></figure><h3>Thrid Step:</h3><p>Ok, now I think there is a bug here in how the server works with the email schema. After some testing, this is the case from which I understood what happened. I sent a new request with two different emails [new_user_two@gmail.com, new_email_three@gmail.com]. The server sent the verification to new_email_one@gmail.com.</p><p>Ok, I caught this — the server accepts the array but treats it as a string instead of recognizing it as an array. When it tries to send the email, it finds that the string (array) is not a valid email to send the verification token to, so it sends it to the last valid email associated with this user ID, which is new_email_one@gmail.com, because it is a valid email schema and the token can be sent to it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/400/1*GtWyTawZN2DbZ4Jxq-sYVA.gif" /></figure><p>Now we’ve found a bug — let’s create a scenario.</p><h3>Fourth Step:</h3><p>I thought a lot about the scenario; I will describe it here as I wrote it in the report. I think there is wrong validation in the email field — that’s why I tried to change my email to a victim’s email with a small change so the server would fail to send verification to the intended email and instead send it to the last valid one. So I tried again. Let’s say we have three emails for testing:</p><ol><li>Current email</li><li>Attacker email</li><li>Victim email</li><li>I go to change email and change email from current_email@gmail.com → attacker_email@gmail.com, intercept the request, send it to Repeater, and forward it. A new token is sent to attacker_email@gmail.com.</li><li>I then request to change attacker_email@gmail.com to victim_email@gmail.com, but if I leave it like that the server responds that the email is already used.</li><li>So I try to add a space or null byte at the end to make it look identical to the victim email, but it does not work. I then try to add \u0000 at the end but the server removes it. After a lot of searching I find \u00A0, which refers to a non‑breaking space So the eamil look like victim_email@gmail.com\u00A0. The server accepts it and the email looks in the response like victim_email@gmail.com (with a trailing space), and the verification token is sent to attacker_email@gmail.com.</li><li>After I receive the OTP and confirm it, Now i could act as a Victim ,I can directly access the account and change the name to the victim’s name and set their photo. Now I am the victim. There are many ways to abuse this: for example, sending invitations will appear to come from victim_email@gmail.com (e.g., “<a href="/service/mailto:victim@gmail.com">victim@gmail.com</a> has invited you to collaborate”), and other site features that display the email could be affected.</li></ol><h4><strong>Note:</strong></h4><p>Now the database stores two emails that look the same but are not identical: the real victim email victim_email@gmail.com and a fake email victim_email@gmail.com with an extra trailing space.</p><p>After that i Reprot it and after talk alot with triager and program thay take it to that you the impact come from it lead to social engineering scenarios after more than month they finally accept it a reward me with bounty</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_Js-EV0exd25CfQYkpGAVg.jpeg" /></figure><p>🤍في الاخر اتمني تكون استفدت باي حاجه و متنساش تدعي ل اخواتنا المسلمين و كل بقاع الارض</p><p><a href="http://www.linkedin.com/in/mahmoud-kroush-2a4011289">www.linkedin.com/in/mahmoud-kroush-2a4011289</a><br><a href="https://x.com/MahmoudKroush33">https://x.com/MahmoudKroush33</a><br>“سبحانك اللهم وبحمدك، أشهد أن لا إله إلا أنت، أستغفرك وأتوب إليك”.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=76c1d83868dc" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[XSS and CSRF — A Full Kill Chain]]></title>
            <link>https://medium.com/@kroush333/xss-and-csrf-a-full-kill-chain-4606e435dacb?source=rss-e61e334e1ece------2</link>
            <guid isPermaLink="false">https://medium.com/p/4606e435dacb</guid>
            <category><![CDATA[xss-attack]]></category>
            <category><![CDATA[bug-bounty]]></category>
            <category><![CDATA[xss-vulnerability]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <dc:creator><![CDATA[MahmoudKroush (xgoon)]]></dc:creator>
            <pubDate>Fri, 25 Jul 2025 11:27:31 GMT</pubDate>
            <atom:updated>2025-11-19T18:58:15.419Z</atom:updated>
            <content:encoded><![CDATA[<h3>XSS and CSRF — A Full Kill Chain</h3><p>بسم الله، والصلاة والسلام على رسول الله , الحمد لله الذي عَلَّمَ بالقلم.. عَلَّمَ الإنسانَ ما لم يَعْلَم والصلاةُ والسلامُ على خيرِ مُعَلِّمي الناسِ الخير محمد</p><h3>Who am I</h3><p>I’m <strong>Mahmoud Kroush</strong>, a recently graduated from Faculty of Engineering. I’m aiming to become a penetration tester and currently work part-time as a bug bounty hunter.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AB5HrJRFNy9fhzEP8gD80Q.jpeg" /></figure><h3>Introduction:</h3><p>In this write-up, I will walk you through how I was able to achieve an Stored XSS while working on an external bug bounty program. By chaining together an XSS vulnerability with a CSRF flaw, I successfully executed my attack scenario. You might already be imagining how this could Done, but let’s dive into the details of the site and identify the weak points.</p><h3>Brief About the Target:</h3><p>The target is a large, well-established company that offers a variety of features, including tools for creating and managing documentation, building websites, and more. The company also operates a domain dedicated to educational purposes, providing content on how to use its services, let’s call it <a href="http://freeplastain.education/"><strong>freeplastain.education</strong></a>. This site hosts several blog posts that guide users step by step on how to use each function of the platform. While I was investigating the blog, a particular feature caught my attention — it looked interesting and potentially vulnerable.</p><h3>Finding the XSS</h3><p>The site offers a feature that allows users to bookmark important parts of a blog post while reading. This helps users easily navigate back to key sections using a top navigation bar where the bookmarks are displayed.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/549/1*fpimkzq2-aO4RDqqiei_eA.png" /></figure><p>While testing this feature, I discovered that the bookmark value is stored directly in the page and rendered without proper sanitization. When I tried injecting an XSS payload,</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/531/1*4-EqteLnV3qD97JEl40qzA.png" /></figure><p>it executed and popped up on the page — <strong>coooool!</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/500/1*GN0fwON24fpXTZRf4ycL9w.gif" /></figure><p>Okay, let’s take a look at the request. It was a <strong>POST</strong> request, and the response was returned normally — with my XSS payload reflected in the response.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/850/1*cPJJF7UVidoa4XP0hVqe-A.png" /></figure><p>When you discover an XSS vulnerability in a <strong>POST</strong> request, several attack scenarios may come to mind. One of the first ideas I had was to try changing the request method from <strong>POST</strong> to <strong>GET</strong> — but , that attempt failed. Then I thought: <em>What about a CSRF . ping-pong</em> you’re on the right track. Let’s dive deeper into it!</p><h3>Chaining CSRF</h3><p>When hunting for a CSRF vulnerability, there are a few key questions you should ask in your specific case:</p><ol><li><strong>Is there a CSRF token in the parameters or headers?</strong> — <em>No</em></li><li><strong>Is there a Referer or Origin check?</strong> — <em>No</em></li><li><strong>Is the SameSite cookie flag set in the request?</strong> — <em>No</em></li></ol><p>OK , I can move forward and create a CSRF proof of concept (PoC).</p><p>I used the Burp Suite extension to generate the PoC.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/674/1*7ZQjsXYsJFkN_G2824a8lQ.png" /></figure><h3>Finally</h3><p>I sent the crafted CSRF request and ran it in my browser.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1019/1*QunqphUYYFxN4tZe3EwADw.png" /></figure><p>As expected, the XSS payload executed and popped up and we have full <strong>Scenario</strong> .</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/500/1*IzAy_r03oxlBH0n4jP_nqg.gif" /></figure><p>Mission accomplished!</p><h3>Full Scenario</h3><ol><li>On the target site, I created a new bookmark and intercepted the request using <strong>Burp Suite</strong>, then sent it to the <strong>Repeater</strong> tab.</li><li>I visited <a href="https://xss.report/">https://xss.report/</a> and copied an XSS payload edit it make it suitable. I inserted the payload into the bookmark name parameter, making sure to URL-encode it to keep the request valid.</li><li>Using <strong>Burp’s CSRF PoC Generator</strong>, I created a CSRF proof-of-concept.</li><li>I then sent the PoC link to the victim. Once the victim opened it, the XSS payload executed, and the data (including session, cookies, and other sensitive information) was sent to me via <strong>xss.report</strong>.</li></ol><p>🤍اتمني تكون استفدت باي حاجه و ان شاء الله يكون في احسن الفتره الجايه</p><p>🤍متنساش تدعي لو اخواتنا في غزه<br><a href="http://www.linkedin.com/in/mahmoud-kroush-2a4011289">www.linkedin.com/in/mahmoud-kroush-2a4011289</a><br><a href="https://x.com/MahmoudKroush33">https://x.com/MahmoudKroush33</a></p><p>“سبحانك اللهم وبحمدك، أشهد أن لا إله إلا أنت، أستغفرك وأتوب إليك”.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4606e435dacb" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[From Guest to Admin: Chaining Pre-Account Takeover with Privilege Escalation]]></title>
            <link>https://medium.com/@kroush333/from-guest-to-admin-chaining-pre-account-takeover-with-privilege-escalation-fe970e80cd25?source=rss-e61e334e1ece------2</link>
            <guid isPermaLink="false">https://medium.com/p/fe970e80cd25</guid>
            <category><![CDATA[bug-bounty]]></category>
            <category><![CDATA[bug-bounty-tips]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <dc:creator><![CDATA[MahmoudKroush (xgoon)]]></dc:creator>
            <pubDate>Fri, 16 May 2025 12:56:53 GMT</pubDate>
            <atom:updated>2025-05-16T13:00:24.264Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/961/1*ee55E_1sYimDeFcOifRUOg.png" /></figure><p>بسم الله، والصلاة والسلام على رسول الله<br> الحمد لله الذي عَلَّمَ بالقلم.. عَلَّمَ الإنسانَ ما لم يَعْلَم والصلاةُ والسلامُ على خيرِ مُعَلِّمي الناسِ الخير محمد</p><h3>Who am I</h3><p>I’m <strong>Mahmoud Kroush</strong>, a third-year student at the Faculty of Engineering. I’m aiming to become a penetration tester and currently work part-time as a bug bounty hunter.</p><h3>Introduction:</h3><p>In this write-up, I will walk you through the steps I took to escalate a <strong>Pre-Account Takeover</strong> into a <strong>Privilege Escalation</strong>. The vulnerability was caused by a lack of proper email confirmation in the application’s account creation process.</p><h3>Brief About the Target:</h3><p>The target was a platform designed to create and host web pages using a drag-and-drop interface for images and editable text. The platform supports collaborative work through workSpace that include different user privilege levels:</p><ol><li><strong>Owner</strong> → Has full privileges within the workspace.</li><li><strong>Admin</strong> → Has the same privileges as the Owner, except they cannot control (remove or edit) the Owner.</li><li><strong>Publisher</strong> → Can publish and edit pages, and manage content within the publishing scope.</li><li><strong>Editor</strong> → Can edit existing pages and create new ones, but cannot publish them.</li><li><strong>Guest</strong> → Has view-only access.</li></ol><h3><strong>Frist Step:</strong></h3><p>While exploring the target — let’s refer to it as <strong>freepalestine.com</strong> — I started learning about its different privilege levels and how users interact with the platform.</p><p>When I signed up on the site, I noticed that I could access my account <strong>immediately</strong>, without needing to confirm my email address. Although a confirmation email was sent, it wasn’t required to activate or access the account.</p><p>At that point, I started thinking about the possibility of a <strong>Pre-Account Takeover</strong> vulnerability. I reported this issue about three or four times in the past, but each time it was marked as <strong>Informative</strong>, so I initially moved on and stopped pursuing it.</p><h3>Second Step:</h3><p>While exploring the target as a <strong>Guest</strong> user, I discovered that I could view <strong>invited users</strong> who were still in a <strong>pending state</strong> (i.e., they had been invited but hadn’t completed their registration yet).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/796/1*SdfBN1eH7ePvZkyDO3zyCA.png" /></figure><p>This caught my attention — I realized that this could potentially be abused as part of a <strong>Pre-Account Takeover</strong> scenario. It made me think: <em>What if I could register using the same email that was already invited and gain access with elevated privileges?</em></p><h3><strong>Third Step:</strong></h3><p>Now, let’s think about how I could <strong>accept the invitation</strong>, even though the invitation is only sent to the user’s email. I tried several methods, such as reusing previous invitation links, abusing requests, and experimenting with different techniques — but none of them worked at first.</p><p>While working on this target with my friend <a href="https://medium.com/@yassentaalab51"><strong>Yassen</strong></a>, he discovered a <strong>GraphQL query</strong> that returned detailed information about the <strong>workspace</strong>, including data about invited users. I decided to further explore this query to see if I could extract any useful information that could help in crafting a successful <strong>Pre-Account Takeover</strong> attack.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/903/1*MO5t7C5yZvHyfpSKzSgMrQ.png" /></figure><p>While searching through the GraphQL response, I found a section that included information about <strong>invited users</strong>, and most importantly, it contained the <strong>invitation ID</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/849/1*43OulQu99YvI928fTuORFg.png" /></figure><p>I then went back to analyze the <strong>invitation acceptance process</strong> and discovered that it uses a <strong>GraphQL mutation</strong> which requires the <strong>invitation ID</strong> to complete the process. This was a key finding — with the ID exposed in the response, I now had what I needed to try accepting the invitation myself.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/898/1*CJFLk2waVjPcGQxnPcR2pg.png" /></figure><p>So let’s put it all together — we now have a complete scenario that leads to <strong>Privilege Escalation</strong> on my actual email account.</p><p>To achieve this, I only needed two things:</p><ol><li>An <strong>authorization token</strong>, which I obtained through the <strong>Pre-Account Takeover</strong>.</li><li>An <strong>invitation ID</strong>, which I discovered through <strong>Excessive Data Exposure</strong> in the API (specifically the GraphQL response).</li></ol><p>With both of these elements, I was able to successfully <strong>accept an invitation</strong> intended for another user and gain access to a <strong>workspace with Admin privileges</strong> — all using my own email.</p><p>Now let’s walk through the full scenario step-by-step to view how dangerous this vulnerability really is.</p><h3>Attack Scenario</h3><ol><li>I start as a <strong>Guest</strong> user, waiting for the <strong>Owner</strong> to invite a user with <strong>Admin privileges</strong>.</li><li>I take the <strong>invited email address</strong> and sign up using it (exploiting the lack of email confirmation).</li><li>Using <strong>Burp Suite</strong>, I find the UserSettingsWorkspace <strong>GraphQL query</strong> in the response.</li><li>I extract the <strong>invitation ID</strong> from this response (thanks to <strong>Excessive Data Exposure</strong>).</li><li>I obtain the <strong>invitation acceptance mutation</strong> (either from GraphQL introspection or from a previously captured request).</li><li>I replace the <strong>authorization token</strong> with the token of the newly registered (attacker-controlled) account, and inject the <strong>invitation ID</strong>.</li><li>I now successfully access the <strong>target workspace</strong> and escalate my privileges from <strong>Guest to Admin</strong>.</li><li>Finally, I remove the original invited email (the one I hijacked) and re-invite it using <strong>my own email</strong> to fully take over the Admin role</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/785/1*oHYUUGHL1Z6OHsWgqQ_iww.png" /></figure><p>ولا اكن حاجه حصلت نكمل زي ما احنا</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/418/1*j_3mFpe8Uswkea5fupVKGQ.png" /></figure><p><strong>Final Note:</strong><br> In the end, the report was marked as a <strong>duplicate</strong> — but honestly, I didn’t mind. I wasn’t in it just for the bounty. What mattered most to me was the <strong>bug i want bug bug bug bug </strong>, the process of discovering something impactful, and sharpening my skills.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BFdE2GPqFrSAJU3oG6xM3w.png" /></figure><p>🤍في الاخر اتمني تكون استفدت باي حاجه و متنساش تدعي ل اخواتنا في غزه و كل بقاع الارض<br> <a href="http://www.linkedin.com/in/mahmoud-kroush-2a4011289">www.linkedin.com/in/mahmoud-kroush-2a4011289</a><br> <a href="https://x.com/MahmoudKroush33">https://x.com/MahmoudKroush33</a><br> “سبحانك اللهم وبحمدك، أشهد أن لا إله إلا أنت، أستغفرك وأتوب إليك”.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fe970e80cd25" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Manual Testing for Privilege Escalation and Broken Access Control My (Methodology)]]></title>
            <link>https://medium.com/@kroush333/manual-testing-for-privilege-escalation-and-broken-access-control-my-methodology-a3b9f41b82a2?source=rss-e61e334e1ece------2</link>
            <guid isPermaLink="false">https://medium.com/p/a3b9f41b82a2</guid>
            <category><![CDATA[bug-bounty-tips]]></category>
            <category><![CDATA[bug-bounty]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[penetration-testing]]></category>
            <category><![CDATA[bug-bounty-writeup]]></category>
            <dc:creator><![CDATA[MahmoudKroush (xgoon)]]></dc:creator>
            <pubDate>Sun, 27 Apr 2025 20:13:49 GMT</pubDate>
            <atom:updated>2025-05-17T04:26:58.455Z</atom:updated>
            <content:encoded><![CDATA[<p>بسم الله، والصلاة والسلام على رسول الله<br>الحمد لله الذي عَلَّمَ بالقلم.. عَلَّمَ الإنسانَ ما لم يَعْلَم والصلاةُ والسلامُ على خيرِ مُعَلِّمي الناسِ الخير محمد</p><h3><strong><em>Who am I</em></strong></h3><p>I’m <strong>Mahmoud Kroush</strong>, a third-year student at the Faculty of Engineering. I’m aiming to become a penetration tester and currently work part-time as a bug bounty hunter.</p><h3>Introduction:</h3><p>In this write-up, I will walk you through my methodology for investigating websites, focusing on how I identify vulnerabilities, particularly in the areas of privilege escalation and broken access control (BAC). In this write-up, I will mention only manual testing — no tools, no scripting — just Burp Suite and Notion to take notes</p><h3><strong>Brief About the Target:</strong></h3><p>I worked on a Vulnerability Disclosure Program (VDP) target on HackerOne. For this example, we’ll use the domain <strong>freeplastain.com</strong>. The platform’s core functionality is that when an email is created, it automatically generates an organization. This organization is responsible for managing projects, with varying privilege levels assigned to both the organization and the projects.</p><p><strong>Privilege Levels:</strong></p><ul><li><strong>Privilege Levels Organization:</strong></li><li><strong>Organization User</strong>: Regular user with limited access.</li><li><strong>Contracts Manager</strong>: Responsible for approving and creating contracts.</li><li><strong>Projects Admin</strong>: Can create and manage projects, including project details.</li><li><strong>Organization Admin</strong>: Manages the overall organization, including user roles and permissions.</li></ul><p>These roles can be combined, granting different privilege combinations across the organization and projects.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LwWy9UiqdDOCDpvpI4Ob8g.png" /></figure><p><strong>Project Privilege Levels:</strong></p><ul><li><strong>Member</strong>: Standard user with basic access to the project.</li><li><strong>Assessor</strong>: Can review project details and provide feedback.</li><li><strong>Manager</strong>: Responsible for overseeing the project, including managing resources and tasks.</li><li><strong>Admin</strong>: Has full control over the project, including configuration and user management ( add , remove …etc ).</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VlZx5SkMhCp5yJ5SwxezOg.png" /></figure><h3>Frist Step:</h3><p>Begin by exploring the website thoroughly. Go through each feature and functionality one by one, observing how each part works. As you explore, make sure to document the following:</p><ul><li><strong>Permissions Check</strong>: For each function, note who has the ability to perform specific actions (e.g., creating projects, approving contracts, managing users) and who does not have access.</li><li><strong>Notion Documentation</strong>: Record your findings in Notion, keeping track of which roles have permission for each function, ensuring clarity and organization for further analysis.</li></ul><p><strong>Organization:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WUtth4kuvPR5SsIM2n6hWA.png" /></figure><p>Project:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4NSy5-xTNQg2kOgwDM5EQg.png" /></figure><h3>Second Step:</h3><ul><li><strong>Launch Burp Suite</strong>: Start by running Burp Suite to intercept and analyze the traffic between the browser and the web application.</li><li><strong>Collect Tokens</strong>: Log in with users of different privilege levels (e.g., Organization User, Projects Admin, Organization Admin) and collect their session tokens. Ensure you capture the tokens for each role to observe how the application handles authorization for different users.</li><li><strong>Document Token Details</strong>: Keep a record of each token , noting the user roles and the corresponding tokens. This will be essential when you analyze requests and look for privilege escalation opportunities</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/693/1*tWCQ0mGEBleoxe-B_Ak6Wg.png" /></figure><h3>Find Bug:</h3><ul><li>As I go through each endpoint, I test different privilege levels to identify potential vulnerabilities like Broken Access Control (BAC), Insecure Direct Object References (IDOR), and privilege escalation issues.</li><li>While testing, I identify an endpoint responsible for adding, removing, and modifying project privileges.</li></ul><p><strong>Test Scenario</strong>:</p><p>1- I attempt to reduce the privilege of a user by using a <strong>Member privilege</strong> token. However, the response I receive indicates that I am not authorized to perform this action, confirming that the system enforces access control to some extent at this point.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/884/1*sXONTmC-_5AjT12JNfbRng.png" /></figure><p>2- In the request body, users are identified by a JSON object.</p><p>For example:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/586/1*SHV0O4MXYiPpSmcjgWePyg.png" /></figure><p>3-I then attempted to delete the entire user object from the request body.</p><p>Surprisingly, the server responded with a 200 OK status, indicating the deletion was successful, even though this action was not expected.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1015/1*c13UNINp26u2dOiTemTWbQ.png" /></figure><p>And when i try to login by using the deleted user Find that he not able to access the project</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/277/1*QksK0mXIXb_dRq_mKvZ0Ng.png" /></figure><p>Using the same request, I went back to Burp Suite’s Repeater and sent the request again. the user was re-added to the project.</p><p>Ok let’s go I find bug , After successfully manipulating the user deletion and re-adding process, I decided to escalate my privileges further. I focused on the <strong>permissions array</strong> in the request body.<br>Where I tried to delete permissions but couldn’t, I forgot to try to add it. Ok, let’s try it. I copied the permission array of the Admin, edited mine, and guess what? Yes, I became Admin.</p><p><strong>User 4 Befor :</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*mB5cADT6lxAGJIa0QN1hww.png" /></figure><p><strong>User 4 After :</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1023/1*qQXUHKaBe0PErVMtFULN-w.png" /></figure><p>Let’s goooooooooooooo! It worked! I went and reported it,</p><p>but after two or three days, it was marked as duplicated. This was disappointing =&lt;</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/262/1*e9d2k_GjI0Z6SAauzhv0rQ.png" /></figure><p>🤍اتمني تكون استفدت باي حاجه و ان شاء الله يكون في احسن الفتره الجايه <br> 😂🤍متنساش تدعي لو اخواتنا في غزه و ادعيلي انا كمان اطلع من الدبلكيت بقا</p><p><a href="http://www.linkedin.com/in/mahmoud-kroush-2a4011289">www.linkedin.com/in/mahmoud-kroush-2a4011289</a></p><p><a href="https://x.com/MahmoudKroush33">https://x.com/MahmoudKroush33</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a3b9f41b82a2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Reflected Xss lead me to Hall-of Fame]]></title>
            <link>https://medium.com/@kroush333/reflected-xss-lead-me-to-hall-of-fame-08e6a4c9d500?source=rss-e61e334e1ece------2</link>
            <guid isPermaLink="false">https://medium.com/p/08e6a4c9d500</guid>
            <category><![CDATA[xss-attack]]></category>
            <category><![CDATA[legionhunter]]></category>
            <category><![CDATA[xss-vulnerability]]></category>
            <category><![CDATA[pentesting]]></category>
            <category><![CDATA[bug-bounty-tips]]></category>
            <dc:creator><![CDATA[MahmoudKroush (xgoon)]]></dc:creator>
            <pubDate>Thu, 27 Feb 2025 11:43:28 GMT</pubDate>
            <atom:updated>2025-05-17T04:26:39.510Z</atom:updated>
            <content:encoded><![CDATA[<p>بسم الله، والصلاة والسلام على رسول الله<br>الحمد لله الذي عَلَّمَ بالقلم.. عَلَّمَ الإنسانَ ما لم يَعْلَم والصلاةُ والسلامُ على خيرِ مُعَلِّمي الناسِ الخير محمد</p><h3>Whoami</h3><p>I am Mahmoud Kroush, a student at the Faculty of Engineering, aspiring to become a penetration tester. I also work part-time as a bug bounty hunter.</p><h3>Introduction</h3><p>In this write-up, I will discuss my methodology for finding XSS vulnerabilities. So, without further delay, let’s get started!</p><h4>1️⃣Start by Collecting All Subdomains</h4><p>I use <strong>Subfinder</strong> and <strong>Assetfinder</strong> for subdomain enumeration:</p><pre>subfinder -dL domain.txt -o sub1.txt<br>cat domain.txt | assetfinder -subs-only &gt; sub2.txt</pre><p>Now, I collect all output subdomains into a single file and remove duplicates:</p><pre>cat sub1.txt sub2.txt &gt; subdomains.txt<br>cat subdomains.txt | anew &gt;&gt; final_subdomains.txt</pre><h4>2️⃣ Looking for Alive Subdomains</h4><p>In this step, we target only live subdomains using the <strong>httpx</strong> tool:</p><pre>cat final_subdomains.txt | httpx -o alive_subdomains.txt</pre><h4>3️⃣ Collecting as Many URLs as Possible from the Subdomains</h4><p>In this step, I aim to gather a massive number of URLs to increase my chances of finding vulnerable parameters. I use three tools: <strong>Waybackurls</strong>, <strong>Katana</strong>, and <strong>GAU</strong>.</p><pre>cat alive_subdomains.txt | waybackurls &gt;&gt; urls1.txt<br>cat alive_subdomains.txt | gau &gt;&gt; urls2.txt<br>katana -list alive_subdomains.txt -o urls3.txt</pre><p>then collect all ulrs in one file and fillter it by using uro tool</p><pre>cat urls1.txt urls2.txt urls3.txt &gt;&gt; all_urls.txt<br>cat all_urls.txt | uro &gt;&gt; final_urls.txt</pre><h4>4️⃣ Searching for Specific Endpoints for Testing</h4><p>At this stage, I target endpoints that might be vulnerable to XSS using the <strong>gf</strong> tool:</p><pre>cat final_urls.txt | gf xss &gt;&gt; XSS.txt</pre><h4>5️⃣ Finding Vulnerable Parameters</h4><p>I identify potentially vulnerable parameters using <strong>kxss</strong>, which provides valuable hints about special characters accepted by the parameters:</p><pre>cat XSS.txt | kxss</pre><p>When it worked, I noticed in the results that a parameter called client_id was unfiltered: “client_id Unfiltered: [‘ &lt; &gt; $ | ( ) ` : ; { }]”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iKFf5NsTa_P8cjfNGYRQXA.png" /></figure><h4>6️⃣ Let’s Go to Burp Repeater</h4><p>I open Burp Suite, intercept the request, and send it to the <strong>Repeater</strong>.</p><ul><li>First, I find the reflected point in the response and test with a simple payload:</li></ul><pre>1&lt;script&gt;alert()&lt;/script&gt;</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/469/1*A7QhjaNDzx5G3HO-j5ASqg.png" /></figure><ul><li>I noticed that the closing tag passed with no filtering.</li><li>So, I tried duplicating the payload :</li></ul><pre>1&lt;script&gt;alert()&lt;/script&gt;&lt;script&gt;alert()&lt;/script&gt;</pre><p>Booooooooooooom! It worked</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/649/1*C4usy37vDq3Z_xa0nlVb_Q.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/781/1*XM0yHrbjmG8dcHYQq0q6zA.png" /><figcaption>XSS Executed Successfully</figcaption></figure><p>After confirming the vulnerability, I created a report and submitted it to the program. Shortly after, it was <strong>triaged</strong>, and I was added to the <strong>Hall of Fame on Bugcrowd and the program’s website</strong></p><blockquote><strong>Bugcrowd P2 Severity Vulnerability</strong></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/832/1*KMiSbzdh3EJv9ZpYzGMv4g.png" /><figcaption>Bugcrowd P2 Severity Report</figcaption></figure><blockquote><strong>Bugcrowd Program Hall of Fame</strong></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/402/1*ZTyN05iWe1nlj2xeiUCqDQ.png" /><figcaption>Researcher Bugcrowd Hall of Fame</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/505/1*lfSQug1YoGSQeCnRBrmofA.png" /><figcaption>Researcher Hall of Fame 2025</figcaption></figure><p>I hope this write-up provides you with useful insights!</p><p>“سبحانك اللهم وبحمدك، أشهد أن لا إله إلا أنت، أستغفرك وأتوب إليك”.</p><p>Linkedin: <a href="https://www.linkedin.com/in/mahmoud-kroush-2a4011289/">https://www.linkedin.com/in/mahmoud-kroush-2a4011289/</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=08e6a4c9d500" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>