<?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 Yosef Mostef :) on Medium]]></title>
        <description><![CDATA[Stories by Yosef Mostef :) on Medium]]></description>
        <link>https://medium.com/@yosefmostef99?source=rss-a21460dfdabc------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*5icrDVd_Skno3nf9b5qtVg.jpeg</url>
            <title>Stories by Yosef Mostef :) on Medium</title>
            <link>https://medium.com/@yosefmostef99?source=rss-a21460dfdabc------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Mon, 16 Nov 2026 23:56:49 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@yosefmostef99/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[The Illusion of Instant Block]]></title>
            <link>https://medium.com/@yosefmostef99/the-illusion-of-instant-block-4620220283f9?source=rss-a21460dfdabc------2</link>
            <guid isPermaLink="false">https://medium.com/p/4620220283f9</guid>
            <category><![CDATA[pentesting]]></category>
            <category><![CDATA[information-security]]></category>
            <category><![CDATA[bug-bounty-writeup]]></category>
            <category><![CDATA[bug-bounty]]></category>
            <dc:creator><![CDATA[Yosef Mostef :)]]></dc:creator>
            <pubDate>Fri, 25 Sep 2026 14:16:36 GMT</pubDate>
            <atom:updated>2026-09-25T14:16:36.438Z</atom:updated>
            <content:encoded><![CDATA[<p>بِسْمِ اللَّهِ الرَّحْمٰنِ الرَّحِيمِ، وَالصَّلَاةُ وَالسَّلَامُ عَلَى النَّبِيِّ الْمُجَاهِدِ الشَّهِيدِ…اللَّهُمَّ انصُرْ أَهْلَ غَزَّةَ وَفِلَسْطِينَ وَالسُّودَانَ، وَطَهِّرِ الْأَقْصَى مِنْ دَنَسِ الْيَهُودِ… أَمَّا بَعْدُ</p><p>Hey folks, I’m back after quite a long hiatus… and honestly, it’s good to be back. Oh man, I’d almost forgotten how to write writeups :’(</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/736/1*ld-BvwaYMC7tqLcEYRmKCg.png" /><figcaption>Very sad :’(</figcaption></figure><p>Anyway, enough of that. As usual, on just another random day, I was doing what I love — putting security systems to the test — and I stumbled on a vulnerability built on a familiar idea, just executed differently.</p><p><strong><em>When you get someone deactivated, are you sure that’s really the end of it?</em></strong></p><p>As usual, you test any given feature by first understanding how it works, then trying to break it by any means possible. In this case, the idea was that the site offered a feature letting you deactivate any user — even if that user was an Admin.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bnUHDVk3bEZfZT8uRBRbXA.png" /><figcaption>User</figcaption></figure><blockquote><em>But is that claim actually true?</em></blockquote><p>Take a closer look at the second user’s account, and you’ll find that it was indeed blocked — but the site never invalidated the old sessions. They stayed alive and kept communicating with the server as if nothing had changed.</p><p>I found that the existing session remained valid and retained the account’s pre-deactivation privileges. To prove the vulnerability — and the site made this embarrassingly easy — I changed the account’s name as a simple proof of concept, and even went ahead and blocked the other users, evend delete other admins !!…all from that same “deactivated” account :)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8i-mR9ALeZqkWagqgs9dMw.png" /><figcaption>Change name or anything to proof..</figcaption></figure><p>Digging into it further, the root cause -i think- The application validated the account status during authentication, but failed to revoke existing sessions or enforce the blocked state on subsequent requests.</p><p>The triager was about to close it as Out of Scope, but I walked him through the actual impact :)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/342/1*Hqz7EVnPH0al3weBGGWEYg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/332/1*lHejTaVOJnUlxt-iAtxa6g.png" /></figure><p>And thankfully, it was accepted 😁</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/343/1*neHxaUEGw_T5egIyDE8asQ.png" /><figcaption>الحمدلله</figcaption></figure><p><strong><em>Scenario 2-Trust nothing that comes from the user. And if you do… you’ve already been fooled !</em></strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VMxWzdfBwRAxco2pHOiqrA.png" /><figcaption>The Account of testing</figcaption></figure><p>So I dug a bit deeper, going back through the old requests, and one of them caught my attention. While reproducing the original vulnerability — following the same steps as before: the user gets blocked, and from there we check whether they can still interact with the system, and whether they can still change anything while blocked — I noticed that when I changed the name, the response came back with “status”: “BLOCKED”</p><p>I thought — why not take this parameter and throw it back into the request that edits any profile field?<br> And to my surprise, it actually worked!<br> By manually injecting that parameter into the request body, I was able to modify the status — a textbook Mass Assignment vulnerability.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*l0Q7y_5xKpFoG6lITBo27w.png" /></figure><p>And sure enough, looking at it from the admin’s side, you can actually see that the user lifted the block off themselves.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dgYsWaOzm2lnKLbdDSJtZw.png" /></figure><p>Naturally, the next question was: if “status”was injectable this way, what about “permissions”? I tried the exact same approach — grabbing the permissions field from a legitimate response and replaying it in the profile-update request — but this one was properly validated server-side and got rejected. So the mass assignment issue here was scoped specifically to the status field, not a full privilege-escalation path.</p><p>This isn’t just a passing session-management issue, or some minor edge-case tweak. It strikes at the very structure of your organization on your SaaS platform — a user being able to remove or delete assets, remove members, or block them outright, all while you’re supposed to be confident that the assumption you built your system on can’t be broken.<br> The door was never locked in the first place for it to be “broken.” The door was already wide open.<br> And in other cases, a copy of that very door was sitting in the hands of the person it was supposedly locked against.</p><p>I hope these simple scenarios gave you something useful. And in closing — don’t forget to keep our Uyghur brothers and sisters, the people of Gaza, and oppressed Muslims everywhere in your prayers. Wassalamu alaikum.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/736/1*e1vJ0ughE6xxie1W6qb9Dw.png" /></figure><p>U can find me on:</p><p><a href="http://www.linkedin.com/in/yosef-mostef">www.linkedin.com/in/yosef-mostef</a></p><p><a href="https://t.me/yosefmostef">https://t.me/yosefmostef</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4620220283f9" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Private Polls? Not Really — Exploiting Access Control and Logic Flaws]]></title>
            <link>https://medium.com/@yosefmostef99/private-polls-not-really-exploiting-access-control-and-logic-flaws-943ff0bb5cc4?source=rss-a21460dfdabc------2</link>
            <guid isPermaLink="false">https://medium.com/p/943ff0bb5cc4</guid>
            <category><![CDATA[bug-bounty]]></category>
            <category><![CDATA[infosec]]></category>
            <category><![CDATA[pentesting]]></category>
            <category><![CDATA[information-security]]></category>
            <category><![CDATA[penetration-testing]]></category>
            <dc:creator><![CDATA[Yosef Mostef :)]]></dc:creator>
            <pubDate>Fri, 29 May 2026 17:45:47 GMT</pubDate>
            <atom:updated>2026-05-29T17:45:47.310Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>بسْمِ اللَّهِ الرَّحْمٰنِ الرَّحِيمِ، وَالصَّلَاةُ وَالسَّلَامُ عَلَى النَّبِيِّ الْمُجَاهِدِ الشَّهِيدِ…اللَّهُمَّ انصُرْ أَهْلَ غَزَّةَ وَفِلَسْطِينَ وَالسُّودَانَ، وَطَهِّرِ الْأَقْصَى مِنْ دَنَسِ الْيَهُودِ… أَمَّا بَعْدُ</strong></p><p>One day, while hunting for bugs and going through an intense moment of boredom, curiosity pushed me to keep testing a certain feature on a website. After a lot of trial and error, I eventually discovered not just one issue, but three different vulnerabilities within the same request. Today, I’m sharing a new technical write-up about these three findings:</p><ul><li>A Broken Access Control vulnerability that led to disclosure of voting participants.</li><li>An IDOR vulnerability that allowed unauthorized voting .</li><li>A logic flaw that made it possible to manipulate and cheat the voting system.</li></ul><p>So, grab your favorite cup of coffee, and let’s dive in.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FRrVzUOXldFe8M%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExd3N5dWZyZXBoZWp4c3FnN293Y3pjNXE1ZjNpbjhkMGg3OTk4ZzBueiZlcD12MV9naWZzX3NlYXJjaCZjdD1n%2FRrVzUOXldFe8M%2Fgiphy.gif&amp;image=https%3A%2F%2Fmedia2.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExaWdhaXkzNmhwZ291OGZlbTU0d2t2MnB1bmljemZybzAwYnVyZTY1dyZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2FRrVzUOXldFe8M%2Fgiphy_s.gif&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="326" frameborder="0" scrolling="no"><a href="https://medium.com/media/42b0833e66195c98910246d83abbb81d/href">https://medium.com/media/42b0833e66195c98910246d83abbb81d/href</a></iframe><p>First, let’s understand the nature of the platform we’re dealing with.<br>The platform is primarily designed for publishing personal journals and written posts — something similar to a social media platform, but far more focused on long-form writing and user-generated content.<br>One of its core features was the ability for users to create and publish polls.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/970/1*uqy3R80ctM2MW3j32xl_Ow.png" /><figcaption>Fn</figcaption></figure><p>So, I started by creating a normal post and experimenting with different scenarios until I eventually moved on to testing the poll feature itself. The feature contained multiple configurable options, such as:</p><ul><li>Privacy settings — who can view the poll.</li><li>Voting type — whether a user could vote for both options or only a single one. And several other…</li></ul><p>At that point, I grabbed a piece of paper and started listing assumptions based on the intended business logic of the feature. <strong>The first assumption</strong>: A single user should only be able to cast just one vote.<br><strong>The second assumption</strong>: When privacy settings are configured, only authorized users should be able to access or interact with the poll.<br><strong>The third assumption</strong>:An authorized user should only be able to vote according to the selected poll mode — either:</p><ul><li>Dual-choice voting, where voting for both options is allowed.</li><li>Or single-choice voting, where only one vote should ever be accepted.</li></ul><p>One important note:</p><blockquote><em>I highly recommend interacting with any feature multiple times before testing it deeply. Use it once as a completely normal user to understand its natural workflow, then observe how the requests and responses behave inside Burp Suite. That helps you understand both the application’s logic and the assumptions the developers rely on.</em>And in many cases, those “this should never happen” assumptions are exactly what end up being breakable.</blockquote><p>Now, let’s start breaking those assumptions one by one.The first question was: <em>Can an unauthorized user vote in a poll that is supposed to be private?</em> The answer was a very clear <strong>yes</strong>. To demonstrate this, I first modified the poll’s privacy settings so that the poll would become private and inaccessible to anyone except the post owner.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/801/1*3g9qdMI3-BpCfzEDb0KWRA.png" /></figure><p>Next, I created a second account for testing purposes :) .After creating a poll and submitting a vote, the request looked like this:</p><pre>[ <br>  { <br>  &quot;jsonrpc&quot;: &quot;2.0&quot;, --&gt; JSON-RPC version <br>  &quot;method&quot;: &quot;poll.vote&quot;, ---&gt; purpose of the request <br>      &quot;params&quot;: {<br>           &quot;pollid&quot;: &quot;ID&quot;, --&gt; normal number to determine the poll<br>            &quot;answer&quot;: &quot;1&quot;, ---&gt; the user’s choice <br>             &quot;auth_token&quot;: &quot;ajax:........&quot; ---&gt; token of user<br>             },<br>              &quot;id&quot;: 17 <br>    } <br>]</pre><p>Of course, for any bug bounty hunter, seeing a numeric identifier immediately raises one question: “What happens if I change it? IDOR!!” Naturally, my first instinct was to test whether the endpoint was vulnerable to an IDOR or access control issue. So, I took the same request, replaced the session cookies and the auth_token with those belonging to my second account, and replayed the request against the supposedly private poll. To my surprise, the server responded with HTTP/1.1 200 OK</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*A4nVlsa3Ndd34WjARqjNVA.png" /><figcaption>Wow!</figcaption></figure><p>At that moment, I completely forgot the frustration of more than 11 days without finding any bugs. I became convinced that I had finally discovered a real bug. My confidence increased significantly when I observed that the request was successfully processed and accepted .<br>This was the first vulnerability. However, an important question remained:</p><h4>How could an attacker determine what the original vote values actually were?</h4><p>To answer this, further investigation was required, which led to the discovery of the second vulnerability. While exploring the application, I noticed another endpoint under /_api, which handled different JSON-RPC methods. In a separate request, I identified a method called poll.get &amp; poll.participants.</p><p>poll.get --&gt;This method revealed sensitive poll data, including the results and the number of votes cast. Even more critically, another related method — poll.participants — exposed the list of users who participated in the poll, effectively leaking participant accounts.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Z0EwoLfXcS8x1Y7V_A1Dag.png" /><figcaption>I changed the choice in the request</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5k7MmX9NR-v503GBERpd4Q.png" /><figcaption>The choice with ok!</figcaption></figure><p>At that point, things started to become significantly more interesting. By this stage, two separate vulnerabilities had been identified:</p><ul><li>One allowed full access to poll data, including results and participants.</li><li>The other allowed unauthorized users to cast votes in polls, even when they were marked as private.</li></ul><p>When combined, these bugs created a much more impactful security scenario. An attacker could enumerate polls across the platform -this can be automated easily :) -, And retrieve sensitive information about participants, and still interact with private polls as if they were legitimate users. This effectively broke the core security assumptions of the system: privacy, authorization, and voting integrity. After submission, the report was reviewed and triaged. It was accepted as a valid bug and classified as Medium severity, with a CVSS score of 6.5. Both vulnerabilities were acknowledged and accepted together.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ior3SCDeB3GBgXPKy2TqvA.png" /></figure><p>Unfortunately, the platform only rewards high or critical severity reports. At that point, I started thinking more strategically about how to elevate the impact of the findings. I went back to the original assumptions I had written down earlier and revisited them carefully. One assumption stood out in particular:</p><blockquote><em>A user should only be able to vote as a single entity.</em></blockquote><p>This raised an important question:</p><p>What if this constraint could also be bypassed?</p><p>If that assumption could be broken, the impact would increase significantly. Instead of simply casting one unauthorized vote, an attacker could potentially scale the behavior and vote multiple times, effectively impersonating multiple users — whether 10, 100, or more — depending on how the backend validates identity and vote uniqueness.</p><p>This shifted the focus from simple authorization issues to a deeper logic flaw in how voting integrity is enforced. This turned out to be the third vulnerability in the write-up: a <strong><em>race condition</em></strong>. At this point, I took the private poll vote request and sent it through Burp Suite Repeater, distributing it across multiple tabs and send them parallelly. The result was unexpected but highly impactful. By race condition leading to double voting. The system effectively counted repeated request from the same account as multiple distinct votes. As a result, the final tally became:</p><ul><li><strong>83.3%</strong> for the option selected by the attacker</li><li><strong>16.7%</strong> for the option selected from my second account</li></ul><p>In simple terms, the backend treated each parallel request as if it originated from a different user, allowing the same account to artificially inflate its voting power and manipulate the final poll result.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VZHishzC7o9Cd2-YK1T1jg.png" /></figure><p>Unfortunately, the triage response was the following:</p><blockquote><em>“Hello, the third issue is out of scope since this is vote cheating. This type of issue is specifically excluded from the scope. The illustrated problem does not lead to any financial risks, therefore I have no grounds to explicitly add this one into scope. The report will remain as it is.”</em></blockquote><p>This meant that, within the same request -just change method :) -, three separate vulnerabilities were identified: two of them were accepted and validated -IDOR &amp; BAC-, while the third was rejected as out of scope- race condition-.<br>Even though all three issues existed in the same voting mechanism, only the access control and data leakage vulnerabilities were considered valid for the program, while the vote manipulation issue was excluded from eligibility. I continued researching and testing different methods within the same request, but unfortunately I did not discover anything beyond these three vulnerabilities. The report was ultimately accepted as a Medium severity issue, but it did not qualify for a monetary reward. And i was added as the 11th person to the Hall of Fame.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/323/1*MJ2OURZxkw4EafULTdbayA.png" /><figcaption>الحمدلله</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OHx1EsDSZfO_6aNzowGUVg.png" /><figcaption>Nice</figcaption></figure><p>However, as my first accepted Medium report, it was still a significant milestone for me. It confirmed that the approach worked and that the findings were valid from a security perspective. Since then, my focus has shifted toward consistently chain the impact of vulnerabilities by extracting system assumptions and systematically breaking them :)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/530/1*iSKaoBfIZVXSJeLAytFQKA.png" /></figure><p>To sum it all up: Start by using the feature as a regular user — no testing mindset yet, just normal interaction. Then grab a pen and paper, or whatever you prefer — Notepad, Notion, anything — and write down every assumption you can think of. From there, start thinking about how to break each one. Finally, bring AI into the loop and research similar vulnerabilities that have been found in features like this one. That’s it. See you soon, and happy hacking :)</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=943ff0bb5cc4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[When the UI lies and new powers are discovered.]]></title>
            <link>https://medium.com/@yosefmostef99/when-the-ui-lies-and-new-powers-are-discovered-bb39d97beeaf?source=rss-a21460dfdabc------2</link>
            <guid isPermaLink="false">https://medium.com/p/bb39d97beeaf</guid>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[bugbounty-writeup]]></category>
            <category><![CDATA[bug-bounty]]></category>
            <dc:creator><![CDATA[Yosef Mostef :)]]></dc:creator>
            <pubDate>Wed, 11 Mar 2026 17:27:03 GMT</pubDate>
            <atom:updated>2026-03-11T17:27:03.364Z</atom:updated>
            <content:encoded><![CDATA[<h3>Finding PII via Broken Access Control.</h3><p><strong>سْمِ اللَّهِ الرَّحْمٰنِ الرَّحِيمِ، وَالصَّلَاةُ وَالسَّلَامُ عَلَى النَّبِيِّ الْمُجَاهِدِ الشَّهِيدِ…اللَّهُمَّ انصُرْ أَهْلَ غَزَّةَ وَفِلَسْطِينَ وَالسُّودَانَ، وَطَهِّرِ الْأَقْصَى مِنْ دَنَسِ الْيَهُودِ… أَمَّا بَعْدُ</strong></p><p>Hello folks ,Yosef here .Today I will share the idea of one vulnerability I discovered recently a <strong>PII exposure</strong> across multiple endpoints in organization level ,no custom tools nothing expect my mind and burp.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/736/1*sX-gXXwahTu8wC9v4ir_7g.jpeg" /></figure><p>In my last <a href="https://medium.com/@yosefmostef99/phantom-roles-surviving-removal-with-delete-powers-and-bypassing-add-restrictions-404ea6439ad0?source=user_profile_page---------0-------------a21460dfdabc----------------------">blog post</a>, when I said that I had finally started <strong>hunting seriously</strong> and stopped procrastinating, I meant it. Since then, I have submitted <strong>more than 20 reports</strong> to both large and small companies. Some were closed as <strong>NA by bots</strong> or marked as <strong>Duplicate (Dup)</strong>.</p><p>Of course, I no longer aim to report the <strong>largest possible number of vulnerabilities</strong>. Instead, I focus only on the <strong>better and stronger ones</strong>, and I try to <strong>chain vulnerabilities together</strong> to reach the <strong>highest possible impact</strong>. That has been my strategy.</p><p>I continued hunting throughout the past period, averaging <strong>around 6 hours per day</strong>, trying to stay consistent. One evening, I was discussing things with my brother and my friend <strong>Mohamed</strong> — I recommend following his <a href="https://medium.com/@canitey">account</a>, he is a good person. He suggested that I try another platform. So I started searching until I finally <strong>found what I was looking for</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/314/1*n_oowixB-wTdqQ36n_Gs_A.jpeg" /><figcaption>my reaction when he said this :)</figcaption></figure><p>After a lot of searching and digging, my eyes landed on a platform that was also recommended to me by him.At first, when I started using it, I thought to myself that since it seemed heavily used, it would probably be difficult to find any vulnerabilities in it. But then I told myself: <em>what do you have to lose? Just try — you won’t lose anything.</em></p><p>So I picked a target and started using it normally, just like any regular user. My goal was simply to understand its logic — how it works and how it behaves.</p><p>That’s how the first day ended, after a long session that honestly felt quite boring.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*G-Y8Jl9EMTY_2siyay5cbw.jpeg" /><figcaption>me after this day :(</figcaption></figure><p>On the second day, I started working seriously. I told myself there was no special security system to worry about, put my trust in God, and began.</p><p>The target — let’s call it <strong>target.com</strong> — provides four roles in its system:</p><ul><li><strong>Admin</strong> → Can do anything.</li><li><strong>Manager</strong> → Can do anything except editing users and handling payments.</li><li><strong>User</strong> → Limited access due to the large size of the application.</li><li><strong>Viewer</strong> → Can only view the call center tab and nothing else, and is isolated from other users in the organization.</li></ul><p>I told myself the situation was simple: start by <strong>breaking the “no” rules and the assumptions</strong> that the developers had built into the system.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/736/1*LvpaRCKHUubCW7w1uuR0VQ.jpeg" /></figure><p>So what is <strong>“no”</strong>?<br> <a href="https://www.youtube.com/watch?v=IaUl6g82C1Q">“No” refers to <strong>BAC (Broken Access Control)</strong> or <strong>logic bugs</strong></a> — the kinds of vulnerabilities that come from breaking the restrictions developers assume users will follow.So I started with the first step: <strong>what are the boundaries?</strong>Simple. Whatever is allowed within the interface is the <strong>expected behavior</strong>, and <strong>anything we manage to access outside of that becomes a “no.”</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YWHpRxg6TYCbgI6OCwmoZQ.png" /></figure><p>Every time I tried to reach <strong>admin endpoints</strong> using the <strong>viewer role</strong>, I kept running into a <strong>403 response</strong>.</p><p>I paused for a moment and thought about it more. Then I told myself: <em>sooner or later, I’m going to get you ! .</em></p><p>So I turned on <strong>Burp</strong> and left it running in the background while I continued browsing the application normally. Eventually, I found an interesting endpoint:</p><pre>/modules/pricebook/users/users</pre><p>This endpoint belongs to another feature in the application called <strong>Price Book</strong>, and it returns <strong>all active users</strong> within the organization.</p><p>I took the endpoint and tested it, and the surprise was that <strong>it worked</strong>. It returned <strong>all the organization’s user data</strong>, including <strong>phone numbers and email addresses</strong>.</p><p>That means the endpoint was exposing <strong>PII at the organization level</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AvpzKYUZEbzxOJ9mttcRWg.png" /><figcaption>Nice Catch :)</figcaption></figure><p>The leaked information included:</p><ul><li><strong>Email</strong></li><li><strong>Office phone</strong></li><li><strong>Mobile phone</strong></li><li><strong>User status</strong> — <em>(0 means active, 1 means suspended)</em></li><li><strong>The user’s internal ID in the platform</strong></li></ul><p>Suddenly my <strong>adrenaline spiked</strong>. I told myself this couldn’t be real — there was no way this was actually a vulnerability. I was sure that someone must have already noticed it and reported it before me.</p><p>But thankfully, I was surprised to see that <strong>the report was accepted</strong>. The severity was <strong>low</strong>, yes, but it was still valid — and the <strong>bounty is on the way</strong>, thankfully.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/853/1*wDIH7QlJqbpYE7MHVNdHbg.png" /><figcaption>الحمدلله</figcaption></figure><p>An attacker with the lowest role (Viewer) could enumerate all users in the organization.</p><p>This allows attackers to:<br>- Harvest corporate emails<br>- Perform targeted phishing<br>- Map internal teams<br>- Identify active or suspended users</p><p>I wrote the <strong>PoC</strong>, documented the steps, took screenshots, and submitted the report. Then I kept digging around.</p><p>Eventually, I found <strong>another endpoint</strong> that was leaking all <strong>user emails and IDs</strong>. The amount of data exposed was smaller, but it was still a vulnerability.</p><p>The endpoint was:</p><pre>GET /modules/reporting-rollups/api/users/GetHubUsers HTTP/2</pre><p>This one came from a <strong>different feature</strong> in the application. I reported it as well, but unfortunately, by the time they started reviewing it, it was closed as a <strong>duplicate (Dup)</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/479/1*Z2PlLC3dPe2814HwBSP3tw.png" /><figcaption>nice try :)</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VQ8oAe_YAtxRed-SzQwFUw.png" /><figcaption>Screen From POC :)</figcaption></figure><p><strong>Lessons Learned:</strong></p><ul><li>Always test hidden endpoints.</li><li>Don’t trust UI restrictions.</li><li>Explore other modules.</li></ul><p>Finally, this was a simple write‑up meant to highlight one idea: <strong>look for the “no.”</strong></p><p>If an endpoint appears secure at first glance, that doesn’t mean the road is truly closed. <strong>All roads lead to Rome</strong> — just because one path is blocked doesn’t mean the others are. It only means you need to keep looking for another way.Security assumptions are often where vulnerabilities hide.<br> Developers build systems based on what they believe users <em>won’t do</em> — and that’s exactly where we should start looking.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/474/1*NYOsPIj27lfR019Mq-YwLw.jpeg" /></figure><p>All praise is due to God, who blessed me with His generosity and provision. I hope everyone is doing well.</p><p><strong>Happy hacking.<br> Happy hunting.</strong></p><p>لو استفدت متنساش تدعيلي بالتوفيق والهداية ..والسلام عليكم</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bb39d97beeaf" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Phantom Roles: Surviving Removal with Delete Powers and Bypassing Add Restrictions.]]></title>
            <link>https://medium.com/@yosefmostef99/phantom-roles-surviving-removal-with-delete-powers-and-bypassing-add-restrictions-404ea6439ad0?source=rss-a21460dfdabc------2</link>
            <guid isPermaLink="false">https://medium.com/p/404ea6439ad0</guid>
            <dc:creator><![CDATA[Yosef Mostef :)]]></dc:creator>
            <pubDate>Tue, 06 Jan 2026 16:47:52 GMT</pubDate>
            <atom:updated>2026-01-06T16:47:52.441Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>بِسْمِ اللَّهِ الرَّحْمٰنِ الرَّحِيمِ، وَالصَّلَاةُ وَالسَّلَامُ عَلَى النَّبِيِّ الْمُجَاهِدِ الشَّهِيدِ…اللَّهُمَّ انصُرْ أَهْلَ غَزَّةَ وَفِلَسْطِينَ وَالسُّودَانَ، وَطَهِّرِ الْأَقْصَى مِنْ دَنَسِ الْيَهُودِ… أَمَّا بَعْدُ</strong></p><p><strong>Hello everyone, Yosef here again with two simple privilege escalation scenarios. Grab a cup of coffee and let’s get started :)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/827/1*dWNP3fzpOWLzst3SXDsgOQ.jpeg" /></figure><p>So,<strong> </strong>Bismillah irRahman irRahim<strong>, </strong>let’s begin the story. For some time, I had been hunting at a lower-than-usual pace — sometimes only two hours per week :( — spending the rest of the time casually browsing the internet. One afternoon, I said to myself: why not push a bit harder?<br> So, I selected a target and challenged myself over the course of an entire week to find six vulnerabilities — and I actually did.<br> Let’s start with the first scenario, which represents the simplest privilege escalation you can possibly imagine.</p><blockquote>First Scenario</blockquote><p>After some time and extensive experimentation, an idea came to mind: why not test one last hypothesis? I went ahead and tried it — and indeed, when a user is removed from a certain entity, they still retain their privileges.<br> Hmm… I know this might not be clear yet, so let’s first understand the platform. The website is an educational platform that connects teachers and students. For this experiment, the scenario involved two roles: the Owner and the the Co-Teacher (and other roles, but I focused on these two roles).<br> The platform allows teachers to communicate with each other through a built‑in chat feature. At that point, I thought — why not test this functionality more closely?<br> To begin, I created two accounts and ensured that both accounts were enrolled in the same class.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AuixE2wjKHekXFuKfUllgg.png" /></figure><p>From the first account, I removed the second user — and the operation was completed successfully.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/226/1*SZZtz5IBHEGpxlPTbpeswA.png" /></figure><p>This is secure behavior… right? Yes, it is.The removed user should no longer be able to access anything. Let’s try accessing the class URL again — or, more specifically, append /message/{chat_id}.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/450/1*v4VJXqelmSxKGVhdJIXYtA.png" /></figure><p>I observed that the second account retained access to the chat and could actively interact with it, despite having been removed from the class. More critically, the account was able to delete the entire group without being a member</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tNy2DSxjqhHBqnD34IdKhQ.png" /></figure><p>The root cause of this issue lies in improper server‑side authorization checks. When a user is removed from a class, the application fails to invalidate or re‑verify the user’s permissions for previously authorized resources, specifically the chat functionality. As a result, endpoints such as /message/{chat_id} rely solely on the existence of a valid session or user ID, without confirming the user’s current membership or role within the class. This allows a removed user to retain unauthorized access and perform privileged actions—such as sending messages or deleting the entire group—despite no longer being an authorized participant.</p><p>Unfortunately, this issue was classified as P4 and marked as a duplicate of another report that had already been published since 2023!!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/769/1*6hREdH6AfamOlFWt2PuxxA.png" /><figcaption>sad :|</figcaption></figure><blockquote>Scenario Two: A Simple, Classic Privilege Escalation Case.</blockquote><p>Alright, let’s take a step back for a moment. Do you remember when I mentioned that I was focusing on only two roles — the first being the Co‑Teacher, and the second the Owner?<br> Let’s start by identifying the differences between them.<br> A Co‑Teacher can perform almost all the same actions as the Owner, with one exception: they cannot add new Co‑Teachers or remove existing ones.<br> At that point, I decided to put these roles under the microscope and thought to myself — let’s try breaking the rules.</p><p>I initially tried to remove another user-from co -teacher-, but the application correctly rejected that action — which was expected.<br> So, I moved on to the second vulnerability and asked an important question: can a Co‑Teacher add other Co‑Teachers?<br> To answer this, I first needed to understand how users are added. Using the Owner account, I added a third user. The process was straightforward: the Owner enters the email address → the invited user receives the invitation and completes the process. Perfect.<br> I then took detailed notes on the endpoint, captured the request, and saved it in Notion. Next, while logged in as a Co‑Teacher, I refreshed the page, intercepted the request, and sent it to repeater.<br> From there, I modified several parameters — most notably the endpoint itself — and replaced the email address with that of a completely different, fourth user.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*d_W-Vvi8laPLzUH7CcL5xQ.png" /><figcaption>yeah we made it ✌</figcaption></figure><p>As shown in the attached image, the application accepted the request from the Co‑Teacher — despite the fact that this role is not permitted to add other users — and successfully sent the invitation-<strong>Classic Server-Side Authorization Bypass-</strong>.To further validate the issue, let’s check the email and confirm who actually sent the invitation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_SChySWX1hBL1X1uwrwW6w.png" /><figcaption>From the tested account</figcaption></figure><p>Well, we did it man….we did it. Unfortunately, the journey ended with multiple duplicates. Still, the challenge continues until I land a bounty, so stay tuned for more scenarios to come.</p><p>The warrior within me remains relentless — constantly pushing harder and refusing to stop until the objective is achieved.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/736/1*P0umkWhxldyc3_p8e5wa4w.jpeg" /></figure><p>Keep Hacking and see u in next writeup</p><p>ولا تنسى الدعاء لأخواتنا في فلسطين والسودان والإيغور…والسلام عليكم</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=404ea6439ad0" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Kicked Out, Still Posting: An Authorization Bypass in Class API]]></title>
            <link>https://medium.com/@yosefmostef99/kicked-out-still-posting-an-authorization-bypass-in-class-api-c290c4e8d4d5?source=rss-a21460dfdabc------2</link>
            <guid isPermaLink="false">https://medium.com/p/c290c4e8d4d5</guid>
            <dc:creator><![CDATA[Yosef Mostef :)]]></dc:creator>
            <pubDate>Wed, 29 Oct 2025 10:04:28 GMT</pubDate>
            <atom:updated>2025-10-29T10:04:28.332Z</atom:updated>
            <content:encoded><![CDATA[<h3>Kicked Out, Still Posting: An Authorization Bypass in Class API.</h3><p><strong>بِسْمِ اللَّهِ الرَّحْمٰنِ الرَّحِيمِ، وَالصَّلَاةُ وَالسَّلَامُ عَلَى النَّبِيِّ الْمُجَاهِدِ الشَّهِيدِ…اللَّهُمَّ انصُرْ أَهْلَ غَزَّةَ وَفِلَسْطِينَ وَالسُّودَانَ، وَطَهِّرِ الْأَقْصَى مِنْ دَنَسِ الْيَهُودِ… أَمَّا بَعْدُ</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/262/1*AG874IyNfTlceXUbcYE4SQ.jpeg" /></figure><p>One evening, after returning from an exhausting day at university, I decided to see what I should do. So I thought, why not continue my journey in securing websites? I took the site and began the inspection.</p><p>After a not so short while — about two hours — I discovered a strange and scenario…</p><p>Let’s start by <strong>understanding the website</strong> before anything else — because if you don’t grasp how it fundamentally works, you can’t break it.</p><p>The site is an <strong>educational platform</strong> specialized in virtual classrooms, providing tools for students to facilitate memorization and revision. Let’s assume its name is <strong>target.com</strong>.</p><p>It offers features to <strong>create classrooms</strong>, assign <strong>teachers</strong> and <strong>students</strong>, and manage learning resources.</p><p>So, I decided to <strong>break the rules they assumed were unbreakable</strong>.</p><h3>First: What are the “rules” in the first place?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/965/1*Gzbb0Ca_NPjB6FVWsa5AUQ.png" /><figcaption>Admin rules</figcaption></figure><p>The rules, as shown in the image, are very simple: the owner or admin can allow a student to add or delete a <strong><em>SET</em></strong>— within the classroom, of class, not globally…So, let’s assume a simple scenario where I am the attacker.</p><p>First, what is the <strong><em>unacceptable behavior</em></strong> in this group? It is something like modifying a set you do not own-within the classroom, of class, not globally- and you cannot delete any user because that is outside your permissions. You also cannot add or delete a set if the admin has disabled the add/delete feature for sets. And you cannot invite someone if the admin has disabled invitations.</p><p>Great — so we have several scenarios and operations to break. But today, I will share <strong><em>one operation</em></strong> that I successfully broke — maybe I broke the rest or maybe not… that’s what we’ll find out in the coming days.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/700/1*edBklDjfI8lpabLQVMkAzQ.jpeg" /><figcaption>🤔😁🤷‍♂️</figcaption></figure><p>Let me think out loud with you a bit…</p><p>What if a malicious person wanted to disrupt the group or classroom**? <br>Let’s say the class is for <strong><em>history studies</em></strong>, and I go ahead and add and share content about <strong><em>science </em></strong>or <strong><em>math</em></strong> on the classroom.</p><p>That’s illogical, undesirable, and would annoy the class owners.</p><p>Now… what if I could do this without anyone being able to stop me — at all? <br>Even the <strong><em>classroom owner</em></strong> can’t do anything about it.</p><p>And what if I could pull all this off while <strong><em>not even being in the group anymore</em></strong> — I was in it, but got kicked out because of my behavior…</p><p>Yet, even after being expelled, I still manage to add public sets to the group…Brilliant, right?<br>Is the picture starting to clear up now? 😈</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/480/1*UNdRrobwgMtKmtJHpMPm9A.jpeg" /></figure><p>Yes, the <strong>removed member</strong> was able to add <strong>sets</strong> to the group even if the admin <strong>disabled the set-adding feature</strong> for the group, and the admin <strong>could not stop them at all</strong>.</p><p><strong>Vulnerability Reproduction Steps:</strong></p><ol><li>The attacker accesses the <strong>endpoint</strong> responsible for associating public sets with classrooms.</li><li>Create a <strong>private classroom</strong> of your own and add <strong>any public set</strong> to it.</li><li><strong>Intercept the request</strong> using a proxy (e.g., Burp Suite), capture it, and send it to the <strong>Repeater</strong> tab.</li></ol><pre>POST /webapi/3.2/class-sets/save</pre><p>Request Body:</p><pre>{&quot;data&quot;:[{&quot;classId&quot;:{YOUR_CLASS_ID},&quot;setId&quot;:{PUBLIC_SET_ID}}]}Request Body:</pre><p>Now, <strong>modify the classId</strong> to the <strong>ID of the target group</strong> — the one you were <strong>previously in and got kicked out of</strong>.</p><p><strong>Send the request</strong>…<strong>And… yes — it works!</strong> We successfully exploited the vulnerability.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/567/1*RWmzq33x-O4JeGsYtxIG2Q.jpeg" /></figure><p>the vulnerability:</p><p><strong>Bypass of &quot;Add Set&quot; Disable Feature</strong> Yes – the vulnerability <strong>fully bypasses</strong> the admin&#39;s setting to disable adding sets. Even when the admin explicitly turns off the &quot;Allow students to add sets&quot; option, the attacker can still add public sets via the API. The added sets appear in the classroom <strong>as if nothing happened</strong>, with no indication of unauthorized access….and <strong><em>the admin can never prevent this action!</em></strong></p><p><strong>Classroom ID (classId) Accessibility</strong> The classId is <strong>trivially obtainable</strong> because it is <strong>exposed directly in the classroom URL</strong>.The key issue is whether the backend validates a user’s permission to modify the specified class (classId) on the server side. If authorization is enforced only in the client (UI) and not rechecked by the API, an attacker can tamper with the request body (e.g., change classId) to perform unauthorized actions.</p><p><em>In the end, are you waiting for a happy ending…?</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/556/1*zZCnnOkHDLYpHIbIabE0YQ.jpeg" /></figure><p>You’re a <strong>bug bounty hunter</strong>: you hunt for a vulnerability, write a report, submit it, and move on with your life as if nothing happened.</p><p>If it gets <strong>accepted</strong> — that’s your <strong>rizq</strong> (provision), and you’ve earned it. If it comes back as <strong>NA or Informative</strong>, even though the bug has real <strong>security impact</strong> — it’s <em>still</em> your rizq.</p><p>Just <strong>keep striving</strong>, keep searching, and <strong>trust in your Lord</strong> that you <em>will</em> receive your provision…</p><p><strong>But you won’t get your rizq if you don’t chase it!</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*usneuDEuMBLb_upVaV4pRw.png" /><figcaption>الحمدلله</figcaption></figure><p><strong>Final advice…</strong></p><p>If you find a vulnerability, <strong>try applying the same logic</strong> to other parts of the site — wherever the context allows.</p><p>With high probability, you’ll discover <strong>more instances of the same vulnerability type</strong> in different places.</p><p>Just <strong>keep digging</strong> — you might just find them!</p><p>Happy hacking! 🧑‍💻</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c290c4e8d4d5" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How I Turned a “No” into an IDOR..]]></title>
            <link>https://medium.com/@yosefmostef99/how-i-turned-a-no-into-an-idor-e27383af8709?source=rss-a21460dfdabc------2</link>
            <guid isPermaLink="false">https://medium.com/p/e27383af8709</guid>
            <dc:creator><![CDATA[Yosef Mostef :)]]></dc:creator>
            <pubDate>Fri, 12 Sep 2025 03:47:51 GMT</pubDate>
            <atom:updated>2025-09-12T03:47:51.720Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>بِسْمِ اللَّهِ الرَّحْمٰنِ الرَّحِيمِ، وَالصَّلَاةُ وَالسَّلَامُ عَلَى النَّبِيِّ الْمُجَاهِدِ الشَّهِيدِ… اللَّهُمَّ انْصُرْ أَهْلَ غَزَّةَ وَفِلَسْطِينَ، وَطَهِّرِ الْأَقْصَى مِنْ دَنَسِ الْيَهُودِ… أَمَّا بَعْدُ</strong></p><p><strong>One day, while sipping a cup of my favorite drink — coffee — a thought crossed my mind… Why not try hunting for vulnerabilities again after a break that lasted for about two months…? So I said to myself, why not? Let’s go!</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/228/1*GwuDl1qK62AVCOMNk_tjMg.jpeg" /></figure><p>I decided to look for a target on my favorite platform, “Bugcrowd,” and after a while I identified a specific target and decided to work on it — let’s assume its name is <em>anything.com</em>. First of all, before hunting for any vulnerabilities, you need to understand the application. So, I opened Burp Suite and spent around five minutes actively working on the site, and the rest of the hour analyzing the requests and how the site handled them. After a few hours — I can’t recall exactly how many — I found an interesting endpoint:</p><pre>https://anything.com/webapi/3.9/users/{id}</pre><p>Yes, yes… so let’s try it, why not…</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/474/1*SLXKvxhzHQC0H1BoRETTHA.jpeg" /></figure><p><strong>IDOR (Insecure Direct Object Reference) is a vulnerability where an attacker manipulates a parameter or an object identifier (such as a user ID, file ID, or record number) in a request. If the application does not properly enforce authorization checks, the server may return data or perform actions that belong to another user. For example, by changing the </strong><strong>id=123 in a request to </strong><strong>id=124, an attacker might gain access to another user’s profile, sensitive information, or even modify someone else’s data. This happens because the system relies only on the provided identifier without validating whether the requesting user is authorized to access the requested object.</strong></p><p>So, I changed it to another ID — an account I had created for testing the vulnerability and, of course, to confirm it — and the server responded with unauthorized access….</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/736/1*u8DwzbdlRJ76c1UdziIHSw.jpeg" /></figure><p>Yes, yes, this is something very likely to happen, but wait… do you see what I see?<br><em>The request was originally using the </em><em>PUT method, so I thought—why not try changing it to </em><em>GET to see if the response would be different? To my surprise, it actually returned information from another user’s account.</em>…Did we really do it? Did we actually find an IDOR?<br> But wait — the fields returned this time were fewer.</p><p>In the first response, the server disclosed highly sensitive information such as the user’s email, date of birth, and more. However, in this new response, although it didn’t reveal equally critical fields, it still exposed sensitive information.</p><p>After some digging, I noticed it returned fields such as:</p><ul><li>lastModified: showing the exact date/time when the profile was last updated.</li><li>timezone: which should only be accessible to the account owner.</li><li>subscriptionPlan: indicating whether the account is on a free or paid plan.</li><li><em>Most importantly, the </em><em>timestamp field, which can have multiple implications. In many systems, it may be used in token generation, digital signatures, session validation, or other security-critical mechanisms.</em></li></ul><p>Okay, I’ll go ahead and submit the report to the site. Honestly, at first, I wasn’t sure whether this information was sensitive or not, so I thought, why not? There’s nothing to lose. I submitted it, and the thought crossed my mind that it might be a P2 worth $1,500 — the dream of getting rich quick.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/630/1*JO4VMvSs0k_frcAZ8TAEaA.jpeg" /></figure><p><strong>However, the platform had a different opinion. After six days, it was classified as a P3 and also marked as a duplicate.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ksqQwr2PS0YPPj3bflWiHQ.jpeg" /></figure><p>And at that moment, I remembered a quote I read a year ago:<br> <em>“Maybe I didn’t succeed in anything this evening, but that’s okay — I will try again tomorrow.”</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/626/1*pspNDe0X0cNSoA2VC2eAbw.jpeg" /></figure><h3>Lessons Learned:</h3><ul><li>Always test different HTTP methods — sometimes a simple change can reveal hidden data.</li><li><strong>Not all bugs are hard to hunt; sometimes just changing one thing can make the difference.</strong></li><li>Going through the experience matters, even if your intuition turns out to be wrong.</li></ul><p>And finally…<strong>Happy hacking!</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e27383af8709" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Cryptology Fundamentals Exam-CAPT-]]></title>
            <link>https://medium.com/@yosefmostef99/cryptology-fundamentals-exam-capt-72c1dfbe83df?source=rss-a21460dfdabc------2</link>
            <guid isPermaLink="false">https://medium.com/p/72c1dfbe83df</guid>
            <dc:creator><![CDATA[Yosef Mostef :)]]></dc:creator>
            <pubDate>Sun, 24 Aug 2025 11:27:29 GMT</pubDate>
            <atom:updated>2025-08-24T11:27:29.174Z</atom:updated>
            <content:encoded><![CDATA[<blockquote><strong>بِسْمِ الَّذِي عَلَّمَ بِالْقَلَمِ، عَلَّمَ الإِنسَانَ مَا لَمْ يَعْلَمْ، وَالصَّلَاةُ وَالسَّلَامُ عَلَى خَيْرِ مُعَلِّمِ النَّاسِ، عّلّى خّيرِ مُحِمد… أَمَّا بَعْدُ….</strong></blockquote><p><strong>I’m really excited to share my very first article on Medium! ⚔:-D<br> It will cover a set of questions from the <em>Cryptology Fundamentals</em> section of the Certified Associate Penetration Tester (CAPT) certification.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jaSLUMrjWWLldJCHNBpZOw.png" /></figure><h3>Key Points Summary 😁</h3><h3>1. Introduction</h3><ul><li>Cryptology = <strong>Cryptography</strong> (protecting data) + <strong>Cryptanalysis</strong> (breaking/enhancing security).</li><li>History: from Ancient Egypt → Caesar cipher → Enigma → modern cryptography.</li><li>Importance: ensures <strong>CIA (Confidentiality, Integrity, Availability)</strong> for sensitive data (e.g., government info, certificates).</li></ul><h3>2. Cryptography Basics</h3><ul><li><strong>Encryption</strong> = algorithm + key; <strong>Decryption</strong> = reverse process.</li><li>Keys ensure correct operation.</li><li><strong>Public / Private Keys</strong> in asymmetric encryption.</li><li><strong>Algorithms</strong>: AES, RSA, DES.</li><li><strong>Protocols</strong>: SSL/TLS, SSH, IPsec.</li></ul><h3>3. Cryptography vs. Cryptanalysis</h3><ul><li><strong>Cryptography</strong>: protects data using math.</li><li><strong>Cryptanalysis</strong>: attempts to break protection or find weaknesses.</li><li>Types: <strong>Symmetric</strong> (same key) vs. <strong>Asymmetric</strong> (public/private keys).</li><li>Principles: <strong>Kerckhoffs’s Principle</strong> (security relies on key secrecy), <strong>Shannon’s Theory</strong> (ciphertext must look random).</li><li>Attacks: Brute Force, Dictionary, Side-Channel, Statistical.</li></ul><h3>4. Symmetric Encryption</h3><ul><li>Uses one key for both encryption and decryption.</li><li><strong>Block Ciphers</strong>: DES, 3DES, AES.</li><li><strong>Stream Ciphers</strong>: RC4, Salsa20, ChaCha20.</li><li><strong>Modes of Operation</strong>: ECB, CBC, CFB, OFB, CTR.</li><li><strong>Pros</strong>: fast, efficient, simple.</li><li><strong>Cons</strong>: key distribution/management challenges, single key compromise = total loss.</li><li><strong>Applications</strong>: VPNs, SSL/TLS, data storage.</li></ul><h3>5. Asymmetric Encryption</h3><ul><li>Two keys: <strong>public</strong> (shared) + <strong>private</strong> (kept secret).</li><li><strong>Algorithms</strong>: RSA, ECC, ElGamal, DSA.</li><li><strong>Pros</strong>: solves key distribution, stronger authentication, supports digital signatures.</li><li><strong>Cons</strong>: slower, requires more resources.</li><li><strong>PKI</strong>: system for managing public/private keys and digital certificates.</li><li><strong>Applications</strong>: SSL/TLS, secure email, VPN, authentication, certificates.</li></ul><h3>6. Hash Functions</h3><ul><li>Convert input of any size → fixed-length value.</li><li><strong>Properties</strong>: one-way, collision resistance, avalanche effect.</li><li><strong>Algorithms</strong>: MD5, SHA-1, SHA-2, SHA-3, RIPEMD-160.</li><li><strong>Attacks</strong>: rainbow tables, collision attacks.</li><li><strong>Uses</strong>: password storage, signatures, data integrity verification.</li></ul><h3>7. Digital Signatures</h3><ul><li>Use asymmetric cryptography for: <strong>authentication, integrity, non-repudiation</strong>.</li><li><strong>Process</strong>: hash document → encrypt hash with private key → receiver verifies with public key.</li><li><strong>Algorithms</strong>: RSA, DSA, ECDSA.</li><li><strong>Applications</strong>: email security, contracts, financial transactions, e-government.</li></ul><h3>8. Cryptographic Protocols</h3><ul><li>Rules and algorithms for secure communication, identity verification, data integrity.</li><li><strong>Examples</strong>: SSL/TLS, IPsec, SSH, PGP, Kerberos, OAuth.</li><li><strong>Applications</strong>: web security, network security, email, authentication, data protection.</li></ul><h3>9. Cryptanalysis Techniques</h3><ul><li>Science of breaking encryption and testing security.</li><li><strong>Techniques</strong>: Brute Force, Dictionary, Side-Channel, Known Plaintext, Frequency Analysis, Differential, Linear, Man-in-the-Middle.</li><li><strong>Applications</strong>: security evaluation, digital forensics, military/intelligence, cyberattacks.</li></ul><h3>Finally, the exam — along with its solution</h3><p><em>First Question</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/628/1*zzKfL4uh-k-1QUVfTTpe9A.png" /><figcaption>I managed to get this one easily because the lecturer had already explained the definition at the start</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/638/1*dIN8D-6mmT5eVXoDxQR4rw.png" /><figcaption>in this question, we were given multiple choices: the first option described the process of converting plain text into ciphertext, the second was the actual correct answer, the third referred to a process that encrypts data but produces a fixed — though not always consistent — output, and the last one simply meant plain text, not the operation itself. That’s why the right choice was the second one: decryption</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/607/1*6Pya2H__b2tPQCAj56_6BA.png" /><figcaption>This question was about asymmetric encryption. The idea was to identify something that everyone can access in this algorithm. The first option was invalid because it referred to the private key, which is only held by the owner and never shared publicly. The second option described decryption, which clearly wasn’t relevant since the question asked about what is owned, not the process of converting ciphertext back to plaintext. The third option was plain text — which honestly made no sense in this context. So naturally, the last option was the correct one</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/608/1*jg62vGBwXJCSXHiopZwDYQ.png" /><figcaption>This was a tricky yet well-crafted question. At first glance, all the options might look equally valid, but the correct answer is the first one. That’s because non-repudiation is considered a secondary principle, not a primary goal of cryptography</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/592/1*6ZzpUlRfQwUqJBnnpkhcrw.png" /><figcaption>Easy</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/626/1*y-by0gRUEVxVzvVOdwDHfA.png" /><figcaption>Asymmetric</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/555/1*s-x2u0oM5Re794Btl_Y7gQ.png" /><figcaption>Base64 is Encoding!!!</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/601/1*cRnbUaj68kGOm8KbsJVrjg.png" /><figcaption>This question is brilliant because it really tests your research skills. Even if you don’t look it up, you could still figure it out through experimentation — which is exactly what I did (admittedly, out of a bit of laziness) :)</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/503/1*QkrtQwQbey9gbf8LAKASBA.png" /><figcaption>This is a straightforward question that mainly tests your knowledge of protocols. The one that isn’t encrypted is Telnet. According to Wikipedia, <em>‘</em><strong><em>Telnet transmits all information, including usernames and passwords, in plaintext, so it is not recommended for security-sensitive applications such as remote management of routers. Its use has largely declined in favor of SSH, although some extensions were proposed to add encryption.’-</em></strong><a href="https://en.wikipedia.org/wiki/Telnet">https://en.wikipedia.org/wiki/Telnet</a>-<br> So, the correct answer here is Telnet.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/613/1*VJtQmf4MgNheiNg0LSgJew.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/625/1*C1K_OCVano6BWEkL2duSPg.png" /><figcaption>I was only able to solve this question by using two steps: first, I identified the type of hash using <a href="https://hashes.com/en/tools/hash_identifier">https://hashes.com/en/tools/hash_identifier</a>,Once I knew the algorithm, I entered it into <a href="https://md5hashing.net/hash/">https://md5hashing.net/hash/</a></figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/637/1*tN0pTejp36-J9HHdc-kGmw.png" /></figure><p>And with that, we’ve reached the end of the article. Wishing you a great day — take care and goodbye. 🥰</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=72c1dfbe83df" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>