<?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 Suhel Kathi on Medium]]></title>
        <description><![CDATA[Stories by Suhel Kathi on Medium]]></description>
        <link>https://medium.com/@suhelkathi?source=rss-c7f6f5369f58------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*6j1XC0a1hQx9gsG4</url>
            <title>Stories by Suhel Kathi on Medium</title>
            <link>https://medium.com/@suhelkathi?source=rss-c7f6f5369f58------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 11:10:01 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@suhelkathi/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[From Foothold to Domain Admin — The Real Attacker Journey Inside Active Directory]]></title>
            <link>https://suhelkathi.medium.com/from-foothold-to-domain-admin-the-real-attacker-journey-inside-active-directory-6f1057d0ef5b?source=rss-c7f6f5369f58------2</link>
            <guid isPermaLink="false">https://medium.com/p/6f1057d0ef5b</guid>
            <category><![CDATA[penetration-testing]]></category>
            <category><![CDATA[domain-controller]]></category>
            <category><![CDATA[active-directory-attack]]></category>
            <category><![CDATA[privilege-escalation]]></category>
            <category><![CDATA[active-directory]]></category>
            <dc:creator><![CDATA[Suhel Kathi]]></dc:creator>
            <pubDate>Mon, 22 Dec 2025 15:28:06 GMT</pubDate>
            <atom:updated>2026-01-19T14:08:18.349Z</atom:updated>
            <content:encoded><![CDATA[<h3>From Foothold to Domain Admin — The Real Attacker Journey Inside Active Directory</h3><p>Active Directory is rarely breached in one dramatic move. Real-world compromises happen through a chain of small, logical steps. Once an attacker gains a foothold inside the network, even with a low-privileged account, Active Directory becomes a map waiting to be explored.</p><p>This post walks through how attackers <em>actually</em> approach AD after initial access — how they think, what they look for, and how a single user account can realistically lead to full domain control.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/635/1*OpIJUUhZ3PhQ3ErPpHLJZw.png" /></figure><h3>Stage 1: Initial Foothold — The First Crack</h3><p>Attackers usually enter through the easiest door, not the strongest one. Phishing emails, malicious attachments, spoofed login pages, leaked credentials, or compromised third-party services are common entry points.</p><p>At this stage, the account obtained is typically:</p><ul><li>A normal employee user</li><li>No admin rights</li><li>Limited access to systems</li></ul><p>That’s enough. In AD, visibility matters more than privilege early on.</p><p>The objective here is simple: <strong>stay alive and stay unnoticed</strong>.</p><h3>Stage 2: Local Discovery — Understanding</h3><p>Before touching the domain, attackers study the compromised host. This machine is the launchpad.</p><p>They quietly check:</p><ul><li>Who is currently logged in</li><li>Whether admins or service accounts have logged in before</li><li>What access the user already has locally</li></ul><p>A workstation used by IT staff is far more valuable than a regular user laptop. Even without admin rights, attackers now start planning the domain-wide view.</p><h3>Stage 3: Domain Enumeration — Mapping AD Like an Attacker</h3><p>Once inside the network, the attacker’s mindset shifts. Active Directory is no longer abstract — it’s a graph of relationships.</p><p>The main goal here is <strong>situational awareness</strong>.</p><p>Attackers want answers to questions like:</p><ul><li>Who are the users?</li><li>Which groups exist?</li><li>Who are the domain admins?</li><li>What service accounts are running?</li><li>Where are the domain controllers?</li><li>What trusts and permissions connect everything?</li></ul><p>Tools such as <strong>BloodHound</strong>, <strong>PowerView</strong>, and <strong>Impacket</strong> are commonly used at this stage. They don’t exploit vulnerabilities — they <strong>read what AD already exposes</strong>.</p><p>BloodHound, in particular, turns directory data into attack paths, highlighting misconfigurations, dangerous privileges, and the shortest route to high-value targets.</p><p>By the end of enumeration, attackers clearly know:</p><ul><li>Which accounts matter</li><li>Which systems are valuable</li><li>Which misconfigurations can be abused</li></ul><p>At this point, the attack becomes targeted instead of exploratory.</p><h3>Stage 4: Credential Access — Turning Access into Power</h3><p>With a clear map, attackers shift focus to credentials.</p><p>Passwords, hashes, and Kerberos tickets are now the priority.</p><p>Common attacker interests include:</p><ul><li>Cached credentials on endpoints</li><li>NTLM hashes for reuse</li><li>Kerberos tickets already present in memory</li><li>Service account credentials</li></ul><p>If local admin access is obtained anywhere, credential extraction becomes trivial. Even without plaintext passwords, hashes and tickets are often enough to move laterally.</p><p>This is where attackers start gaining real momentum.</p><h3>Stage 5: Kerberos Abuse — Trust Becomes the Weapon</h3><p>Kerberos itself is secure — attackers abuse <strong>how it’s deployed</strong>.</p><p>Service accounts with SPNs are a goldmine. Attackers request service tickets and extract them for offline cracking, hoping for weak or reused passwords. Tools like <strong>Rubeus</strong> and <strong>Impacket</strong> make this efficient and quiet.</p><p>Other times, attackers simply reuse stolen Kerberos tickets to impersonate users on different machines. No password required, no alerts triggered.</p><p>The key shift here:<br> <strong>Attackers stop breaking security and start using it correctly.</strong></p><h3>Stage 6: Lateral Movement — Expanding the Blast Radius</h3><p>Now comes controlled movement across the network.</p><p>Using legitimate Windows protocols, attackers hop between systems:</p><ul><li>File servers</li><li>Application servers</li><li>Database servers</li><li>Jump hosts</li></ul><p>This movement blends in with normal admin activity. The goal is to land on machines where privileged users authenticate, increasing the chances of capturing higher-value credentials.</p><p>Noise is avoided. Patience wins.</p><h3>Stage 7: Privilege Escalation — Finding the Shortcut</h3><p>At this point, attackers already know their best path forward.</p><p>They exploit:</p><ul><li>Weak ACLs</li><li>Over-permissioned groups</li><li>Delegation misconfigurations</li><li>Trusted but poorly secured servers (backup, monitoring, management)</li></ul><p>BloodHound often confirms these paths before action is taken, reducing risk.</p><p>One misconfigured object is enough.</p><h3>Stage 8: Domain Admin — Game Over</h3><p>Once domain admin or equivalent replication rights are obtained, control is complete.</p><p>Attackers can now:</p><ul><li>Dump all domain credentials</li><li>Modify Group Policy</li><li>Create stealthy persistence</li><li>Disable security tools</li><li>Deploy ransomware or exfiltrate data at scale</li></ul><p>This didn’t happen because of one bug — it happened because of <strong>many small design and hygiene failures</strong>.</p><h3>Why This Matters</h3><p>Active Directory attacks are not magic. They are predictable, repeatable, and heavily automated.</p><p>Understanding this journey helps:</p><ul><li>Pentesters model realistic attack paths</li><li>Defenders detect early warning signals</li><li>Organizations remove shortcuts before attackers use them</li></ul><p>If AD is the backbone of the enterprise, then <strong>attack paths are the pressure points</strong>.</p><blockquote>📢 Disclaimer:<br>This blog is for <strong>educational use only</strong>. Never test or exploit third-party services without permission.</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6f1057d0ef5b" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How Attackers See Active Directory — Internal vs External Threat Perspectives]]></title>
            <link>https://suhelkathi.medium.com/how-attackers-see-active-directory-internal-vs-external-threat-perspectives-8d3ae96b58e7?source=rss-c7f6f5369f58------2</link>
            <guid isPermaLink="false">https://medium.com/p/8d3ae96b58e7</guid>
            <category><![CDATA[active-directory]]></category>
            <category><![CDATA[active-directory-attack]]></category>
            <category><![CDATA[active-directory-security]]></category>
            <category><![CDATA[penetration-testing]]></category>
            <category><![CDATA[insider-threat]]></category>
            <dc:creator><![CDATA[Suhel Kathi]]></dc:creator>
            <pubDate>Sat, 01 Nov 2025 19:01:56 GMT</pubDate>
            <atom:updated>2026-01-29T14:27:45.266Z</atom:updated>
            <content:encoded><![CDATA[<h3>How Attackers See Active Directory — Internal vs External Threat Perspectives</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FjBd4R-A9MNagFmPQ0dlUQ.png" /></figure><p>Active Directory is the crown jewel for many organizations. To an attacker, AD is not just a directory of users — it’s a map, a toolkit, and the fastest route to control. But the way an attacker approaches AD depends on where they start: are they outside the network trying to phish their way in, or are they already inside as a low-privilege user or rogue insider? The view, tactics, and priorities change — and defenders who understand both viewpoints win.</p><p>This post breaks down how attackers think, step-by-step, from <strong>external</strong> and <strong>insider</strong> vantage points, what they hunt for, and what defenders should watch for first.</p><h3>Two attacker personas</h3><p><strong>1) External attacker (outsider / remote)</strong><br> Starts outside the organization. Typical entry paths:</p><ul><li>Phishing → stolen creds / initial shell.</li><li>Exploiting exposed services (RDP, VPN, web apps).</li><li>Compromised third-party (supply-chain).</li></ul><p>Goal: get in, then escalate, move laterally, and reach domain admin.</p><p><strong>2) Internal attacker (insider / resident)</strong><br> Already has a foothold inside the network — could be a compromised employee, contractor, or a malicious insider with legitimate access.</p><p>Goal: use existing trust to get more privileges silently (less noisy, often more effective).</p><p>Both types share many techniques later in the chain, but their reconnaissance, tools, and stealth tactics differ.</p><h3>Reconnaissance: how attackers map AD</h3><p>Whether external or internal, the first thing an attacker does is build a mental map.</p><p>External recon (before or immediately after breach):</p><ul><li>Public info: LinkedIn, job postings, GitHub, subdomains, Shodan.</li><li>Phishing targets: employees with admin-sounding roles.</li><li>Exposed services: look for RDP, VPN portals, exposed LDAP/LDAPS.</li></ul><p>Internal recon (once inside):</p><ul><li>LDAP queries to list users, groups, OUs.</li><li>SMB enumerations to find shares and SYSVOL.</li><li>DNS and AD replication endpoints to find domain controllers.</li><li>Tools: PowerView, AdFind, BloodHound for relationships.</li></ul><p>Key idea: reconnaissance transforms AD from a concept into a target map of accounts, trusts, services, and weak links.</p><h3>External attacker playbook (typical flow)</h3><ol><li><strong>Initial access</strong><br> Phish an employee, exploit exposed web/VPN, or use leaked credentials.</li><li><strong>Establish foothold</strong><br> Get a shell, persist (scheduled task, malicious service), and avoid detection.</li><li><strong>Internal discovery</strong><br> Run LDAP/SMB scans, enumerate SPNs, gather service account names and potential Kerberos targets.</li><li><strong>Credential theft &amp; escalation</strong><br> Dump LSASS, collect cached hashes, kerberoast service accounts, AS-REP roast weak users.</li><li><strong>Lateral movement</strong><br> Use Pass-the-Hash, Pass-the-Ticket, RDP/SMB hops to reach higher-value hosts.</li><li><strong>Domination</strong><br> DCSync or DCShadow (or simply obtain domain admin creds) → extract secrets, add persistence, modify GPOs.</li><li><strong>Cleanup &amp; persistence</strong><br> Create stealthy admin accounts, schedule covert tasks, hide in backups or service accounts.</li></ol><p>External attackers often generate more noisy telemetry early (phishing, new remote sessions) — but their goal is always to minimize noise once inside.</p><h3>Internal attacker playbook (typical flow)</h3><ol><li><strong>Leverage legitimate access</strong><br> The insider already has credentials and may have access to internal tooling and knowledge of business processes.</li><li><strong>Quiet discovery</strong><br> Use allowed LDAP queries, read AD groups, inspect local resources, check for weakly protected service accounts.</li><li><strong>Exploit misconfigurations</strong><br> Find unmanaged service accounts, weak ACLs, or misapplied delegation that allow privilege changes without triggering external defenses.</li><li><strong>Stealthy escalation</strong><br> Use internal protocols correctly (Kerberos/NTLM) to request tickets or reuse credentials — often without triggering external IPS/IDS.</li><li><strong>Reach for high-value objects</strong><br> Access backups, internal admin tools, or use AD-authenticated services (like SQL servers) to exfiltrate data.</li></ol><p>Insiders can be quiet and surgical. They already bypass many perimeter controls — detection has to rely on behavioral baselines and internal monitoring.</p><h3>What attackers look for first (priority list)</h3><ol><li><strong>Service accounts &amp; SPNs</strong> — prime Kerberoast targets.</li><li><strong>Accounts without Kerberos pre-auth</strong> — AS-REP roast candidates.</li><li><strong>Local admins on many hosts</strong> — credential reuse potential.</li><li><strong>Exposed RDP / VPN</strong> — direct access ramp.</li><li><strong>Weak ACLs / excessive group memberships</strong> — fast privilege jumps.</li><li><strong>Open SMB / SYSVOL</strong> — GPO modifications and script deployment.</li><li><strong>Unmonitored endpoints (OT, backup servers)</strong> — persistence islands.</li></ol><p>This priority list is what attackers automate with scripts — pentesters emulate it, blue teams detect it.</p><h3>Tools of the trade (attacker/pentester toolbox)</h3><ul><li>Recon: <strong>Recon-ng, Shodan, Amass, OSINT tools</strong>.</li><li>Internal AD mapping: <strong>PowerView, BloodHound, AdFind</strong>.</li><li>Kerberos/NTLM manipulation: <strong>Rubeus, Impacket, Mimikatz</strong>.</li><li>Lateral movement: <strong>PsExec, WinRM, RDP, SMB client tools</strong>.</li></ul><p>Note: Tools are neutral — defenders must understand them to detect misuse.</p><h3>Detection &amp; defender playbook — how to spot both attacker types</h3><p><strong>Detecting external attackers</strong></p><ul><li>Monitor for unusual VPN or RDP logins (new devices, odd hours, geolocation jumps)</li><li>Watch for mass mailbox or credential reset requests after phishing campaigns.</li><li>Detect brute force or credential stuffing patterns.</li></ul><p><strong>Detecting insiders</strong></p><ul><li>Behavioral analytics (UEBA): unusual queries to LDAP, atypical access to sensitive shares, unexpected process interactions with LSASS.</li><li>Monitor for unusual use of admin tools from non-admin hosts.</li><li>Alert on new SPN registrations, SYSVOL changes, or sudden group membership changes.</li></ul><p>High-value detection signals:</p><ul><li>Many Kerberos TGS requests for multiple SPNs (possible kerberoast activity).</li><li>AS-REP requests for accounts that rarely authenticate.</li><li>LSASS memory access on endpoints (credential dumping).</li><li>New accounts added to privileged groups or sudden GPO/SYSVOL changes.</li></ul><h3>Hardening &amp; quick wins</h3><p>For defenders who want high impact, low friction changes:</p><ol><li><strong>Inventory &amp; visibility</strong>: maintain an up-to-date inventory of service accounts, SPNs, and exposed endpoints.</li><li><strong>Strong service account hygiene</strong>: use gMSAs, complex rotated passwords, and remove unnecessary SPNs.</li><li><strong>Limit external exposure</strong>: never expose LDAP/SMB/RDP to the internet; require VPN + MFA.</li><li><strong>Detect early</strong>: log and centralize Kerberos/LDAP/Security events to SIEM, tune alerts for mass TGS/AS-REP patterns.</li><li><strong>Least privilege &amp; ACL review</strong>: periodically audit ACLs, remove GenericAll or write rights from non-admins.</li><li><strong>Block common credential capture</strong>: disable LLMNR/NBT-NS, enforce SMB signing, protect LSASS (EDR/credential guard).</li></ol><p><strong>📢 Disclaimer:</strong><br>This blog is for <strong>educational purposes only</strong>. Do not attempt to exploit any application or API you do not own or have explicit permission to test. Always follow ethical hacking guidelines and responsible disclosure practices.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8d3ae96b58e7" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How Authentication Works in Active Directory — Kerberos, NTLM & Real Attack Paths]]></title>
            <link>https://suhelkathi.medium.com/how-authentication-works-in-active-directory-kerberos-ntlm-real-attack-paths-bd9bc6afc105?source=rss-c7f6f5369f58------2</link>
            <guid isPermaLink="false">https://medium.com/p/bd9bc6afc105</guid>
            <category><![CDATA[kerberos]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[information-security]]></category>
            <category><![CDATA[ntlm]]></category>
            <category><![CDATA[active-directory]]></category>
            <dc:creator><![CDATA[Suhel Kathi]]></dc:creator>
            <pubDate>Tue, 21 Oct 2025 12:35:29 GMT</pubDate>
            <atom:updated>2026-01-29T14:27:25.255Z</atom:updated>
            <content:encoded><![CDATA[<h3>How Authentication Works in Active Directory — Kerberos, NTLM &amp; Real Attack Paths</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mt8IlcJ2LVi4IG6xCI0TVw.png" /></figure><p>Active Directory’s most sensitive job is simple in description but critical in impact: <strong>verify who you are</strong>, and then let you access what you’re allowed to. If authentication breaks, everything breaks. If an attacker wins authentication, they often win the network.</p><p>This post explains how Windows authentication works in AD (Kerberos and NTLM), shows the common abuse techniques pentesters use, and ends with practical detection and defence tips you can apply right away.</p><h3>Why authentication is the backbone of AD</h3><p>Think of your company like a gated office building. Authentication is the security desk that checks badges and hands out day passes. A compromised pass means someone can walk into restricted floors. In AD, that “pass” can be a password, a hash, or — in Kerberos — a ticket. Attackers who master these primitives can move from user to admin without ever deploying noisy malware.</p><h3>Two protocols: NTLM (legacy) and Kerberos (default)</h3><p>Windows historically used NTLM; modern AD uses Kerberos by default. Both are still around in real networks, and both matter.</p><p>NTLM</p><ul><li>Older, challenge-response protocol.</li><li>Uses password hashes (NTLM hashes) for authentication.</li><li>Still used in legacy apps, some services, and fallback scenarios.</li><li>Commonly abused via relay and pass-the-hash techniques.</li></ul><p>Kerberos</p><ul><li>Ticket-based protocol (more secure and efficient).</li><li>Uses a trusted Key Distribution Center (KDC) on the Domain Controller.</li><li>Uses two main ticket types (TGT and Service Ticket).</li><li>Best practice for AD — but misconfigurations and SPN handling create real attack opportunities.</li></ul><h3>Kerberos: the ticket flow (simple, practical)</h3><p>When you log in, Kerberos gives you a reusable identity token so you don’t need to keep sending your password.</p><ol><li><strong>Logon → Request TGT</strong><br> You authenticate to the KDC (Domain Controller) with credentials. If correct, the KDC issues a <strong>TGT (Ticket Granting Ticket)</strong>. Think of it as a day pass.</li><li><strong>Access a service → Request Service Ticket</strong><br> For each service (file share, SQL, HTTP), your computer presents the TGT to the KDC and requests a <strong>Service Ticket (TGS)</strong> for that specific service (identified by an SPN — Service Principal Name).</li><li><strong>Present Service Ticket to the service</strong><br> You give the service the service ticket; the service validates it and grants access according to its own permissions.</li></ol><p>Key points:</p><ul><li><strong>TGT</strong> proves identity to the KDC.</li><li><strong>Service Ticket</strong> proves identity to the target service.</li><li>Services still enforce authorization; the ticket proves who you are, not what you can do.</li></ul><h3>Example (real):</h3><p>Alice logs in and gets a TGT. She needs to query sql-prod.suhel.local. Her workstation requests a service ticket for MSSQLSvc/sql-prod.suhel.local. The service ticket is issued and presented to MSSQL; MSSQL then checks if Alice is in the DBAdmin group.</p><h3>Kerberos attack primitives pentesters love</h3><ul><li><strong>Kerberoasting</strong>: attackers request service tickets for accounts with SPNs (typically service accounts). These tickets are encrypted with the service account’s NTLM/kerberos key — attackers extract them and crack them offline to recover service account passwords. Works because many services use weak passwords.</li><li><strong>Pass-the-Ticket</strong>: if an attacker steals a valid service ticket (or TGT), they can inject it into another machine and impersonate the ticket owner.</li><li><strong>AS-REP Roasting</strong>: for accounts that do NOT require Kerberos preauthentication, the KDC returns an encrypted blob that can be cracked offline to reveal the user’s password.</li></ul><p>How pentesters find targets: enumerate SPNs and service accounts, request tickets, and export them for offline cracking. Tools like Rubeus, Impacket (kerberoast scripts), and similar help automate these steps.</p><h3>NTLM: why it still matters</h3><p>NTLM is older but still useful to attackers:</p><ul><li><strong>Pass-the-Hash (PtH)</strong>: attacker steals NTLM hash from a machine and uses it to authenticate without knowing the plaintext password.</li><li><strong>NTLM Relay</strong>: attacker relays an intercepted NTLM authentication to another service (that accepts NTLM), gaining access as the victim. Especially effective against services that don’t require SMB signing or channel binding.</li></ul><p>Legacy apps, misconfigured services, and LLMNR/NBT-NS susceptibility make NTLM relay and PtH practical in many environments.</p><h3>Typical attack chain (how an attacker climbs)</h3><ol><li><strong>Initial foothold</strong>: phishing or exposed service → an account on a workstation.</li><li><strong>Local credential extraction</strong>: dump cached credentials or LSASS memory → get NTLM hashes or Kerberos tickets.</li><li><strong>Lateral movement</strong>: use PtH or pass-the-ticket to access other hosts.</li><li><strong>Kerberoast or AS-REP roast</strong>: obtain service account credentials.</li><li><strong>Privilege escalation</strong>: misuse delegated rights or exploit misconfig to become domain admin (DCSync, DCShadow follow-ups).</li><li><strong>Domain compromise</strong>: with domain admin — deploy GPOs, harvest secrets, persist.</li></ol><h3>Pentester’s quick checklist (what they check first)</h3><ul><li>Are there many SPNs registered to service accounts? (good kerberoast targets)</li><li>Are any accounts set to “Do not require Kerberos preauthentication”? (AS-REP roast targets)</li><li>Is LDAP or LDAPS exposed or misconfigured? (user enumeration)</li><li>Are SMB shares exposed? Is SMB signing disabled? (NTLM relay + lateral tools)</li><li>Is RDP exposed or weakly protected? (direct DC admin access)</li><li>Can I get hashes or tickets from a compromised host? (LSASS, credential caches)</li></ul><h3>Wrap-up — authentication is the AD heartbeat</h3><p>Authentication is the heartbeat of your AD environment. Kerberos and NTLM are the mechanisms that keep that heartbeat steady. Attackers who understand ticketing, SPNs and NTLM flows can escalate from a single workstation to domain takeover — and that’s why defenders must be equally fluent.</p><p><strong>📢 Disclaimer:</strong><br>This blog is for <strong>educational purposes only</strong>. Do not attempt to exploit any application or API you do not own or have explicit permission to test. Always follow ethical hacking guidelines and responsible disclosure practices.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bd9bc6afc105" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Group Policy Objects (GPOs) — The Rulebook of Active Directory]]></title>
            <link>https://suhelkathi.medium.com/group-policy-objects-gpos-the-rulebook-of-active-directory-5bde02f7abe2?source=rss-c7f6f5369f58------2</link>
            <guid isPermaLink="false">https://medium.com/p/5bde02f7abe2</guid>
            <category><![CDATA[penetration-testing]]></category>
            <category><![CDATA[active-directory]]></category>
            <category><![CDATA[windows-security]]></category>
            <category><![CDATA[group-policy-object]]></category>
            <category><![CDATA[active-directory-security]]></category>
            <dc:creator><![CDATA[Suhel Kathi]]></dc:creator>
            <pubDate>Wed, 01 Oct 2025 15:53:34 GMT</pubDate>
            <atom:updated>2025-10-01T15:53:34.216Z</atom:updated>
            <content:encoded><![CDATA[<h3>Group Policy Objects (GPOs) — The Rulebook of Active Directory</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xHOFnQo4hfjv9VEUXMZZ5g.png" /></figure><p>In every organization, rules keep things in order. Whether it’s office timings, dress codes, or security checks at the entrance — without rules, chaos spreads fast. The same applies inside Active Directory (AD). And this is where <strong>Group Policy Objects (GPOs)</strong> come in.</p><p>Think of GPOs as the <strong>rulebook of your digital office</strong>. They define how users and computers behave in the AD environment. From the moment you log in, GPOs are silently shaping what you can or cannot do.</p><h3>A Simple Analogy</h3><p>Imagine a school. The principal sets rules like:</p><ul><li>“Phones must stay silent in class.”</li><li>“Uniform is compulsory.”</li><li>“Internet in the library can only be used for study.”</li></ul><p>Teachers and students don’t set their own rules; they just follow what’s pushed down from above. In AD, the <strong>Domain Controller is the principal</strong>, and GPOs are the rules being distributed to all classrooms (Organizational Units).</p><p>So, when an IT admin wants to make sure USB ports are disabled for all employees, or that every machine locks after 10 minutes of inactivity, they don’t go desk to desk. They simply update the rulebook (GPO), and the change ripples across the whole domain.</p><h3>Real-World Power of GPOs</h3><p>This is where GPOs really shine. They can:</p><ul><li>Enforce password policies and lockout rules.</li><li>Push software updates or restrict app installations.</li><li>Configure firewall settings, disable USBs, or block risky sites.</li><li>Apply different rules to different departments (e.g., IT team gets PowerShell access, Finance team doesn’t).</li></ul><p>All of this happens automatically, without anyone touching individual machines.</p><h3>Why Attackers Care About GPOs</h3><p>Now, here’s the scary part: attackers love GPOs just as much as admins do.</p><p>If an attacker compromises an account with high enough privileges, they can <strong>modify GPOs</strong> to push their own malicious rules across the network. For example:</p><ul><li>Deploy a startup script that downloads malware.</li><li>Disable security software on all machines.</li><li>Create hidden backdoors for persistence.</li></ul><p>With a single malicious GPO update, an attacker doesn’t need to hack every machine — the rules do the job for them. This is why GPOs are often called a <strong>double-edged sword</strong>.</p><h3>Wrapping Up</h3><p>GPOs are the silent guardians of Active Directory. They keep systems organized, secure, and compliant. But like any powerful tool, they demand respect. Proper monitoring, least privilege, and regular audits are the only way to make sure the rulebook isn’t rewritten by the wrong hands.</p><p>So next time your screen locks after 10 minutes or your USB gets blocked, remember: it’s not magic — it’s a GPO quietly doing its job in the background.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5bde02f7abe2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Active Directory Ports & Protocols — The Real Lifelines]]></title>
            <link>https://suhelkathi.medium.com/active-directory-ports-protocols-the-real-lifelines-7e0dd285df80?source=rss-c7f6f5369f58------2</link>
            <guid isPermaLink="false">https://medium.com/p/7e0dd285df80</guid>
            <category><![CDATA[windows-security]]></category>
            <category><![CDATA[pentesting]]></category>
            <category><![CDATA[infosec]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[active-directory]]></category>
            <dc:creator><![CDATA[Suhel Kathi]]></dc:creator>
            <pubDate>Wed, 24 Sep 2025 15:45:35 GMT</pubDate>
            <atom:updated>2025-09-24T15:45:35.020Z</atom:updated>
            <content:encoded><![CDATA[<h3>Active Directory Ports &amp; Protocols — The Real Lifelines</h3><p>In earlier blogs, we covered <strong>what Active Directory (AD) is</strong> and its <strong>architecture</strong>. Now let’s explore something even more critical: the <strong>ports &amp; protocols</strong> that make AD work.</p><p>Think of these as the <strong>lifelines of AD</strong>. Without them, logins won’t work, policies won’t apply, files won’t share — the entire domain will collapse.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gDH0yzdnebM6AyrNMYAcEw.png" /></figure><h3>Kerberos (Port 88) — The Ticket Master</h3><p>Kerberos is the <strong>default authentication protocol</strong> in AD. It works on the idea of <strong>tickets</strong> instead of asking for your password again and again.</p><p><strong>Analogy:</strong></p><ul><li><strong>TGT (Ticket Granting Ticket)</strong> = your <strong>office entry pass</strong>.</li><li>You log in once, DC gives you this pass.</li><li>Valid for a certain time (default 10 hours in Windows).</li><li><strong>Service Ticket</strong> = <strong>canteen/gym/meeting room pass</strong>.</li><li>Every time you access a service (like MSSQL, file server, or printer), Kerberos issues you a specific ticket.</li><li>That service then checks what role you have (Admin vs. Read-only).</li></ul><h3>Types of Kerberos Tickets:</h3><ul><li><strong>TGT (KRB_TGT)</strong> → proves identity to DC.</li></ul><blockquote>This is issued by the <strong>KDC (Key Distribution Center)</strong> inside the Domain Controller.</blockquote><blockquote>It’s like your <strong>office entry card</strong> — proves that you are a valid employee.</blockquote><blockquote>You get it <strong>once when you log in</strong> (after entering username + password).</blockquote><blockquote>Stored on your machine for reuse during that login session.</blockquote><ul><li><strong>Service Ticket (KRB_TGS)</strong> → proves access to a particular service.</li></ul><blockquote>Issued when you want to use a <strong>specific service</strong> (like MSSQL, file share, printer).</blockquote><blockquote>It’s like the <strong>canteen coupon</strong>: you show your entry card (TGT) at the security desk, and they give you a service-specific coupon (TGS).</blockquote><blockquote>Valid only for that service, for a limited time.</blockquote><p><strong>Why attackers love Kerberos:</strong></p><ul><li><strong>Kerberoasting</strong> → stealing service tickets and cracking them offline.</li><li><strong>Pass-the-Ticket</strong> → reusing a valid stolen ticket to impersonate a user.</li></ul><h3>LDAP (Port 389 / 636) — The Librarian of Active Directory</h3><p>LDAP is like the <strong>search engine / librarian</strong> for Active Directory. It doesn’t make rules, it doesn’t enforce access — it just helps you <strong>find information</strong> and <strong>validate credentials</strong>.</p><ul><li><strong>Directory Lookups</strong>:<br> Imagine your company keeps a giant phonebook of every employee, department, and device.</li><li>You ask: <em>“Which groups does Alice belong to?”</em></li><li>LDAP (the librarian) flips through the book and tells you.</li><li><strong>Authentication (LDAP Bind)</strong>:<br> Suppose a service wants to check if Alice’s username/password is valid.</li><li>LDAP compares it against the phonebook.</li><li>If it matches, LDAP says “Yes, Alice is valid.”</li><li>If not, “Invalid credentials.”</li><li><strong>Secure LDAP (LDAPS on Port 636)</strong>:<br> Same librarian, but now sitting inside a <strong>soundproof room</strong> — so outsiders can’t eavesdrop on your questions.</li></ul><p>In short:<br> LDAP = <strong>lookup + validation system</strong>.<br> It makes sure apps, users, and services can quickly find info in AD and confirm if credentials are correct.</p><h3>SMB (Port 445) — The File Room of the Company</h3><p>SMB is like a <strong>shared room</strong> where employees go to pick up files, rules, and tools instead of everyone getting their own copy directly.</p><ul><li><strong>File Sharing</strong>:<br> Imagine you need to give 100 employees access to the same handbook.<br> Instead of printing 100 copies, you place <strong>one copy in the shared room</strong>. Everyone goes there and reads it.</li><li><strong>Group Policy Distribution (GPOs)</strong>:<br> SMB is also how AD delivers company rules (like “disable USBs” or “force strong passwords”).<br> Think of it as the courier service that brings updated policies from IT and pins them on the noticeboard inside the room.</li><li><strong>Replication</strong>:<br> If your company has multiple offices, SMB helps synchronize files/rules so every office’s room has the same content.</li></ul><p>In short:<br> SMB = <strong>the shared file room + courier system</strong>.<br> Whenever you need company files, policies, or updates, you don’t ask IT directly — you just step into the SMB room and pick it up.</p><h3>DNS (Port 53) — The Map Maker</h3><p>Without DNS, AD simply won’t work.</p><p><strong>Analogy:</strong></p><ul><li>Imagine a forest without maps → everyone gets lost.</li><li>DNS is the <strong>map system</strong> of AD, telling clients:</li><li>“Where is the Domain Controller?”</li><li>“Where is the mail server?”</li><li>“What’s the IP of suhel.local?”</li></ul><p>AD relies on <strong>SRV records in DNS</strong> for domain controllers and services.</p><p>If DNS is poisoned or misconfigured:</p><ul><li>Users can’t log in (can’t find DC).</li><li>Traffic may get redirected to malicious servers.</li></ul><h3>RDP (Port 3389) — Remote Watchtower</h3><p>RDP allows admins to <strong>remotely connect to Domain Controllers</strong> or other Windows servers.</p><p><strong>Analogy:</strong><br> It’s like an <strong>elevated watchtower</strong> from where elders (admins) can oversee and manage everything in the forest.</p><p><strong>RDP Risks:</strong></p><ul><li>Attackers brute-force weak credentials.</li><li>Exposed RDP = common entry point for ransomware.</li></ul><h3>Defender’s Checklist</h3><ul><li><strong>Kerberos</strong> → Monitor ticket usage; detect unusual ticket requests.</li><li><strong>LDAP</strong> → Use <strong>LDAPS (636)</strong>; restrict anonymous binds.</li><li><strong>SMB</strong> → Disable <strong>SMBv1</strong>; restrict access to sensitive shares.</li><li><strong>DNS</strong> → Secure DNS servers; enable DNS logging.</li><li><strong>RDP</strong> → Don’t expose directly to the internet; use MFA &amp; VPN.</li></ul><h3>Wrap-Up</h3><p>Active Directory is not just about users and groups — it’s about the <strong>lifelines</strong> that keep the domain alive.</p><ul><li><strong>Kerberos</strong> = Passport system (tickets).</li><li><strong>LDAP</strong> = Phonebook + Doorman.</li><li><strong>SMB</strong> = File Room &amp; Courier.</li><li><strong>DNS</strong> = Map Maker.</li><li><strong>RDP</strong> = Watchtower.</li></ul><p>Understand these ports &amp; protocols → and you understand how AD breathes. Attackers know this very well. Defenders must too.</p><h3><strong>Disclaimer</strong></h3><p>This blog is for <strong>educational use only</strong>. Never test or exploit third-party services without permission.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7e0dd285df80" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Understanding Active Directory Architecture — Explained with Mumbai Local Trains]]></title>
            <link>https://suhelkathi.medium.com/understanding-active-directory-architecture-explained-with-mumbai-local-trains-ff5fd8e60e7c?source=rss-c7f6f5369f58------2</link>
            <guid isPermaLink="false">https://medium.com/p/ff5fd8e60e7c</guid>
            <category><![CDATA[active-directory-attack]]></category>
            <category><![CDATA[active-directory]]></category>
            <category><![CDATA[active-directory-security]]></category>
            <category><![CDATA[red-teaming]]></category>
            <category><![CDATA[penetration-testing]]></category>
            <dc:creator><![CDATA[Suhel Kathi]]></dc:creator>
            <pubDate>Tue, 16 Sep 2025 15:51:58 GMT</pubDate>
            <atom:updated>2025-09-19T18:55:46.998Z</atom:updated>
            <content:encoded><![CDATA[<h3>Understanding Active Directory Architecture — Explained with Mumbai Local Trains</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EAQXHCWbYsmSsZwI678NZg.png" /></figure><p>If you’ve ever worked in IT, you know <strong>Active Directory (AD)</strong> is everywhere. It’s how companies manage users, computers, and resources in one central system. But for beginners, AD architecture can feel… well, abstract.</p><p>So instead of boring theory, let’s explain AD using something every Mumbaikar knows — the <strong>Mumbai Local Train Network</strong> .</p><h3>Step 1: Domain = The City of Mumbai</h3><p>Think of a <strong>Domain</strong> as the city itself.</p><ul><li>In AD, a <strong>domain</strong> is like your company (e.g., suhel.local).</li><li>In Mumbai, the <strong>city boundaries</strong> define who belongs to “Mumbai.”</li></ul><p>Everything inside the city (stations, trains, passengers) is part of the same <strong>domain</strong>.</p><p>So, <strong>Domain = Mumbai city = Your entire company</strong>.</p><h3>Step 2: Domain Controller (DC) = Churchgate Station Master</h3><p>Every system needs authority. In AD, that’s the <strong>Domain Controller (DC)</strong>.</p><p>Now imagine <strong>Churchgate Station Master</strong>:</p><ul><li>They <strong>check tickets</strong> → DC verifies usernames &amp; passwords.</li><li>They <strong>follow schedules &amp; rules</strong> → DC applies company policies (like password rules, login rights).</li><li>They <strong>ensure only valid passengers board</strong> → DC makes sure only authorized users access company systems.</li></ul><p>Without the Station Master, trains would still exist — but anyone could sneak in. Similarly, without a DC, there’s no real control in the domain.</p><p><strong>DC = The boss that keeps the system in order</strong>.</p><h3>Step 3: Organizational Units (OUs) = Suburbs</h3><p>Mumbai has many suburbs: <strong>Borivali, Andheri, Dadar, etc.</strong></p><p>In AD, these are like <strong>Organizational Units (OUs)</strong> — subdivisions inside your company.</p><p>Example:</p><ul><li>Borivali OU → IT Department</li><li>Andheri OU → Finance Department</li><li>Dadar OU → HR Department</li></ul><p>Each suburb (OU) has its own people, rules, and responsibilities. But they’re all still part of Mumbai (the Domain).</p><p><strong>OU = Suburb (department folders inside the company)</strong>.</p><h3>Step 4: Groups = Trains</h3><p>Here’s the fun part.</p><p>Groups are like <strong>trains</strong>. They collect people from different suburbs and move them together for a common purpose.</p><p>Example:</p><ul><li>A <strong>Virar Fast train</strong> = An “IT-Support” Group.</li><li>People from Borivali (IT OU) + Andheri (Finance OU) might both board it.</li><li>A <strong>Churchgate Slow train</strong> = “HR-Payroll” Group.</li><li>Mix of users from Dadar (HR OU) + Borivali (Finance OU).</li></ul><p><strong>Groups bundle people with similar roles/needs — even across different suburbs (OUs).</strong></p><h3>Step 5: Users = Passengers</h3><p>Passengers in Mumbai = <strong>Employees (Users)</strong> in AD.</p><ul><li>Each has a unique ticket → like a <strong>username &amp; password</strong>.</li><li>They belong to a suburb (OU) but can also join trains (Groups).</li></ul><p><strong>Users = Employees traveling across the network.</strong></p><h3>Step 6: Computers = Train Coaches</h3><p>What about the trains themselves? Coaches = <strong>Computers</strong> in AD.</p><p>Just like passengers (users) board coaches, employees use <strong>laptops, desktops, or printers</strong>. All of them are registered inside the same AD system.</p><p><strong>Computers = Devices connected to the company network.</strong></p><h3>Why Attackers Love AD</h3><p>Here’s the scary part.</p><p>Imagine if someone forged a <strong>master railway pass</strong>. They could travel anywhere, board any train, and access any station.</p><p>That’s what happens when attackers compromise the <strong>Domain Controller (DC)</strong>.</p><ul><li>It’s “one key to the kingdom.”</li><li>A single weakness can escalate into <strong>total control of the company network</strong>.</li></ul><p>This is why defenders secure AD tightly… and why pentesters/red teamers love targeting it.</p><h3>Closing Thoughts</h3><p>Active Directory isn’t as mysterious as it sounds.</p><ul><li><strong>Domain = Mumbai city</strong></li><li><strong>Domain Controller = Churchgate Station Master</strong></li><li><strong>OUs = Suburbs (Borivali, Andheri, Dadar)</strong></li><li><strong>Groups = Trains</strong></li><li><strong>Users = Passengers</strong></li><li><strong>Computers = Coaches</strong></li></ul><p>Once you see AD as a city + railway system, the architecture starts to feel natural. And like Mumbai locals, it works best when everything runs on time, with the rules enforced.</p><p>Next up in this journey → we’ll go deeper into <strong>ports, protocols, and how AD actually communicates inside the network</strong>.</p><h3>⚠️ Disclaimer: This explanation is for educational purposes only.</h3><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ff5fd8e60e7c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Active Directory (AD) Basics]]></title>
            <link>https://suhelkathi.medium.com/active-directory-ad-basics-6b3bcaaf3f20?source=rss-c7f6f5369f58------2</link>
            <guid isPermaLink="false">https://medium.com/p/6b3bcaaf3f20</guid>
            <category><![CDATA[penetration-testing]]></category>
            <category><![CDATA[active-directory]]></category>
            <category><![CDATA[active-directory-basics]]></category>
            <category><![CDATA[active-directory-security]]></category>
            <category><![CDATA[active-directory-attack]]></category>
            <dc:creator><![CDATA[Suhel Kathi]]></dc:creator>
            <pubDate>Fri, 12 Sep 2025 13:46:41 GMT</pubDate>
            <atom:updated>2025-10-01T15:16:46.662Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7VYluIb0kWAKZKscVIaTpA.png" /></figure><p>If you’ve ever worked in a company, you know how messy it can get managing hundreds (or thousands) of employees, their computers, and their permissions.</p><h3>What is Active Directory?</h3><p>Think of AD as a <strong>giant company phonebook + security guard + policy manager</strong> rolled into one.</p><ul><li><strong>Phonebook</strong> → It knows every employee, every computer, every printer in the company.</li><li><strong>Security guard</strong> → It decides <em>who</em> can enter <em>what door</em> (login to which system, access which folder).</li><li><strong>Policy manager</strong> → It makes sure rules are applied to everyone’s devices (like disabling USB drives, forcing password changes, or giving software licenses).</li></ul><p>Without AD, every computer in the office would be its own world → you’d have to manage logins, permissions, and rules separately on each. Total chaos.<br> With AD, everything is connected, organized, and controlled centrally.</p><h3>The Tree Analogy</h3><p><strong>Think of AD like a tree:</strong></p><ul><li><strong>Domain (suhel.local)</strong> = The whole tree (your company)</li><li><strong>Domain Controller (DC)</strong> = The root (the boss that keeps the tree alive and in order)</li><li><strong>Organizational Units (OUs)</strong> = Big branches (departments like IT, Finance, HR)</li><li><strong>Groups</strong> = Sub-branches (teams inside departments like IT-Security, IT-Infra, IT-Support</li><li><strong>Users</strong> = Leaves (employees)</li><li><strong>Computers</strong> = Fruits (laptops, desktops, printers connected to the tree)</li></ul><p>The <strong>Domain Controller</strong> (DC) is king here. If attackers compromise the DC, they can control the entire tree.</p><h3>OU vs Groups</h3><p>Here’s the easiest way to understand it:</p><blockquote><strong>OU (Organizational Unit)</strong> = <strong>Folders in a cupboard</strong></blockquote><ul><li>Example: A cupboard with folders named <em>IT</em>, <em>Finance</em>, <em>HR</em>.</li><li>You put employees (user accounts) and their computers inside the right folder.</li><li>OUs are for <strong>organization + rules</strong> (like “USB disabled for Finance laptops”).</li><li>A user can live in <strong>only one OU</strong>.</li></ul><blockquote><strong>Group</strong> = <strong>WhatsApp Groups</strong></blockquote><ul><li>The same employee can join many groups at once.</li><li>Groups are for <strong>permissions</strong> (like VPN access, file share access, software access).</li><li>Groups don’t organize people; they just <strong>control what you can do</strong>.</li></ul><blockquote><strong>Quick Memory Trick:</strong><br> OU = <strong>Folder</strong> (structure + rules)<br> Group = <strong>WhatsApp Group</strong> (access + permissions)</blockquote><h3>Why Companies Love AD</h3><ul><li><strong>Centralized control:</strong> Manage thousands of users/computers from one place.</li><li><strong>Security policies:</strong> Apply rules across all employees (disable USB, enforce password policies, auto-install software).</li><li><strong>Easy resource sharing:</strong> Share printers, drives, and apps with just a few clicks.</li><li><strong>Licensing control:</strong> Manage Office365, Teams, and other licenses centrally.</li></ul><h3>Why Attackers Love AD</h3><p>AD is like the <strong>master key to the kingdom</strong>.</p><ul><li>Compromise one weak user → escalate → eventually compromise the DC.</li><li>Once DC is owned, the attacker owns <strong>everything</strong> (users, groups, computers, policies).</li><li>That’s why AD attacks are so popular in pentesting and real-world breaches (think ransomware).</li></ul><h3>Closing Thoughts</h3><p>Active Directory is both <strong>powerful</strong> and <strong>dangerous</strong>. It makes life easier for companies, but also provides a huge attack surface for hackers.</p><ul><li>As a <strong>developer/IT admin</strong>, think of AD as your safety net but keep it tight.</li><li>As a <strong>pentester</strong>, think of AD as the ultimate target — because once you’re in, the game changes.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6b3bcaaf3f20" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Beyond the OWASP API Top 10 — Injection, Logging Gaps, and Business Logic Flaws]]></title>
            <link>https://suhelkathi.medium.com/beyond-the-owasp-api-top-10-injection-logging-gaps-and-business-logic-flaws-14ecd8d15d29?source=rss-c7f6f5369f58------2</link>
            <guid isPermaLink="false">https://medium.com/p/14ecd8d15d29</guid>
            <category><![CDATA[penetration-testing]]></category>
            <category><![CDATA[owasp-api-security-top-10]]></category>
            <category><![CDATA[api-security]]></category>
            <category><![CDATA[owasp-top-10]]></category>
            <category><![CDATA[injection]]></category>
            <dc:creator><![CDATA[Suhel Kathi]]></dc:creator>
            <pubDate>Sun, 07 Sep 2025 13:56:49 GMT</pubDate>
            <atom:updated>2025-09-07T13:56:49.023Z</atom:updated>
            <content:encoded><![CDATA[<h3>Beyond the OWASP API Top 10 — Injection, Logging Gaps, and Business Logic Flaws</h3><p>Most people know the OWASP API Top 10. It’s like the “greatest hits” of API security risks. But if you have done pentesting or bug hunting, you will agree — attackers don’t stop at those 10. There are plenty of nasty flaws that don’t make the list but show up again and again in real-world APIs.</p><p>Three common favorites for attackers (or nightmares, depending on the day) are: <strong>Injection</strong>, <strong>Insufficient Logging &amp; Monitoring</strong>, and <strong>Business Logic Flaws</strong>. These aren’t new. They’ve been around forever. But they’re still devastating when left unchecked.</p><h3>Injection — the classic that never dies</h3><p>It’s been years, and injection is still here. APIs just gave it a new playground.</p><p>When an API takes input and passes it straight to a database or system without cleaning it, attackers can sneak in commands. The system doesn’t know the difference between “data” and “instructions.”</p><p>Here is a simple test you might run on a vulnerable API:</p><pre>{<br>  &quot;Fname&quot;: &quot;hAPI&quot;,<br>  &quot;Lname&quot;: &quot;Hacker&quot;,<br>  &quot;Address&quot;: &quot;&#39; OR 1=0--&quot;<br>}</pre><p>If the backend SQL database chokes on this and spits back an error like:</p><pre>Error: You have an error in your SQL syntax...</pre><p>That is confirmation — the input was not sanitized, and SQL injection is alive and well.</p><p>And it is not just SQL. APIs are also prone to NoSQL injection, OS command injection, and XML injection. Different flavor, same root problem: trusting user input.</p><h3>Logging &amp; Monitoring</h3><p>Here is the thing: most attacks don’t look like explosions. They look like… noise. Lots of small, repeated requests, often hidden in plain sight.</p><p>If your API isn’t logging or monitoring properly, you won’t even know it is under attack.</p><p>For Example: An API where an attacker could send 10,000 password reset token attempt and not a single log entry was generated. Imagine how easy brute force becomes when nobody is watching.</p><p>Without logs, you lose two things:</p><ul><li><strong>Detection</strong> → You don’t know an attack is happening.</li><li><strong>Forensics</strong> → You can’t figure out what happened after the fact.</li></ul><p>Logging is your early warning system. Without it, you’re flying blind.</p><h3>Business Logic Flaws</h3><p>These are the sneaky ones. Business Logic Flaws don’t exploit bad code — they exploit <strong>bad assumptions</strong>.</p><p>Developers often trust that users will “play nice.” But attackers don’t.</p><p>Here are two cases:</p><ul><li><strong>MFA bypass</strong> — Login API sends MFA=true. An attacker changes it to MFA=false. If the backend doesn’t enforce MFA, the attacker walks right in.</li><li><strong>Partner API abuse</strong> (real case: Experian leak in 2021) — A partner app exposed Experian’s credit check API without restrictions. Anyone could intercept a request, swap the data, and pull <strong>anyone’s credit score</strong>.</li></ul><p>Business logic flaws are everywhere — from refund APIs you can call twice, to inventory APIs you can abuse to scalp PS5s. These are the kinds of bugs scanners rarely catch, but pentesters love.</p><h3>Closing thoughts</h3><p>The OWASP Top 10 is a great starting point, but it is not the whole story. Injection, logging gaps, and business logic flaws live “beyond the list” — and they are often what real attackers go after.</p><p>As a developer: assume nothing, validate everything.<br>As a tester: look for what’s not obvious. That’s usually where the gold lies.</p><p>APIs are only getting bigger and more critical. The attackers know it. We should too.</p><h3>Disclaimer</h3><p>This blog is for educational purposes only. Always responsibly test and handle third-party integrations.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=14ecd8d15d29" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[API10:2023 — Unsafe Consumption of APIs]]></title>
            <link>https://suhelkathi.medium.com/api10-2023-unsafe-consumption-of-apis-08dcec9e9ee0?source=rss-c7f6f5369f58------2</link>
            <guid isPermaLink="false">https://medium.com/p/08dcec9e9ee0</guid>
            <category><![CDATA[api-penetration-testing]]></category>
            <category><![CDATA[owasp-api-security-top-10]]></category>
            <category><![CDATA[api]]></category>
            <category><![CDATA[web-penetration-testing]]></category>
            <category><![CDATA[api-security]]></category>
            <dc:creator><![CDATA[Suhel Kathi]]></dc:creator>
            <pubDate>Wed, 27 Aug 2025 07:54:44 GMT</pubDate>
            <atom:updated>2025-08-27T07:54:44.209Z</atom:updated>
            <content:encoded><![CDATA[<h3>API10:2023 — Unsafe Consumption of APIs</h3><h3>Intro</h3><p>What if I told you that sometimes the real risk isn’t what your server exposes, but <strong>what it trusts</strong>? That’s what <strong>Unsafe Consumption of APIs</strong> is all about. Instead of being the provider, your application becomes the consumer — and if you’re not careful, that third-party API could become the backdoor to your own system.</p><p>In short: treat data from external APIs like user input — with <strong>zero trust</strong>. Because once the third-party provider is compromised, your trusted integration might quietly become your biggest weakness.</p><h3>Why This Happens</h3><p>Many developers treat third-party API data like it’s “clean” — just because it’s not user input. We often skip proper validation, over-rely on TLS without verifying supplemental integrity, and don’t sandbox outputs — the perfect recipe for a backend supply chain attack.</p><h3>Developer-Friendly Scenario</h3><p>Imagine your app fetches weather data via a third-party API:</p><pre>resp = requests.get(f&quot;https://weather-service.com/data/{city}&quot;)<br>return resp.json()</pre><p>One day, that API gets breached and returns this JSON trick:</p><pre>{<br>  &quot;temperature&quot;: 20,<br>  &quot;conditions&quot;: &quot;&lt;/script&gt;&lt;script&gt;stealCookies()&lt;/script&gt;&quot;<br>}</pre><p>Your app passes it downstream. Boom — it becomes a <strong>script injection (XSS)</strong> opportunity in your front-end.</p><p>This isn’t theory — it’s exactly why OWASP warns: <strong>third-party API data = untrusted data</strong>.</p><h3>Summary</h3><p>Unsafe API consumption happens when developers trust external data more than user input. That’s dangerous because attackers don’t need to breach <em>you</em> — they breach your upstream dependencies, and your app becomes the vehicle. It’s a supply-chain attack, implemented quietly and invisibly.</p><h4><strong>Preventative Measures</strong></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/749/1*9FqwcCkmbemki-p1eUQypg.png" /></figure><h3>Final Thoughts</h3><p>When consuming third-party APIs, your trust should be the most cautious security policy. Because if that downstream service fails, your application becomes the attacker’s gateway. Manage and protect your external dependencies as vigilantly as your own backend.</p><p>📢 <strong>Disclaimer:</strong><br> This blog is for <strong>educational use only</strong>. Never test or exploit third-party services without permission.</p><p>If you’re serious about API security, check out <a href="https://www.apisecuniversity.com/"><strong>APIsec University</strong></a> — packed with hands-on labs and free resources.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=08dcec9e9ee0" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[API09:2023 — Improper Inventory Management]]></title>
            <link>https://suhelkathi.medium.com/api09-2023-improper-inventory-management-109a6267049f?source=rss-c7f6f5369f58------2</link>
            <guid isPermaLink="false">https://medium.com/p/109a6267049f</guid>
            <category><![CDATA[owasp-api-security-top-10]]></category>
            <category><![CDATA[api]]></category>
            <category><![CDATA[api-penetration-testing]]></category>
            <category><![CDATA[api-security]]></category>
            <category><![CDATA[web-app-pentesting]]></category>
            <dc:creator><![CDATA[Suhel Kathi]]></dc:creator>
            <pubDate>Wed, 20 Aug 2025 11:43:39 GMT</pubDate>
            <atom:updated>2025-08-20T11:43:39.473Z</atom:updated>
            <content:encoded><![CDATA[<h3>API09:2023 — Improper Inventory Management</h3><h3>Intro</h3><p>APIs evolve fast. Old versions get replaced, test environments pop up, and beta features sneak into production. But here’s the problem: many organizations forget to <strong>retire or secure outdated APIs</strong>.</p><p>This is where <strong>Improper Inventory Management</strong> comes in. It’s when older or non-production APIs remain exposed, often with weaker security. Attackers love this because old endpoints = old vulnerabilities.</p><h3>Example:</h3><p>A pentester is hired to test an e-commerce API.<br>The docs say the latest version is /v3/. But out of curiosity, they try:</p><pre>GET /api/v1/users/123</pre><p>To their surprise, /v1/ is still working. Worse: it returns <strong>full user details</strong> (email, address, phone, hashed password) that /v3/ had stopped exposing.</p><p>No exploit needed — just checking old endpoints. Developers forgot to shut it down, and now sensitive data leaks are possible.</p><h3>How Attackers Exploit This</h3><p>Attackers don’t need zero-days — they just try:</p><ul><li><strong>Old versions</strong>: /api/v1/ or /api/v2/ might still respond, even if /api/v3/ is official.</li><li><strong>Non-production APIs</strong>:</li><li>api.test.example.com</li><li>beta.api.example.com</li><li>/api/private or /api/demo</li><li><strong>Fuzzing &amp; brute force</strong>: Simply guessing endpoints like /uat/ or /staging/.</li><li><strong>Using outdated docs</strong>: Public changelogs, GitHub commits, or archived docs can expose forgotten paths.</li></ul><h3>Why It’s Dangerous</h3><p>Improper inventory management often leads to:</p><ul><li>Sensitive data exposure</li><li>Exploitation of <strong>unpatched vulnerabilities</strong> in old APIs</li><li>Bypass of modern authentication &amp; authorization</li><li>Full system compromise</li></ul><p>Think of it this way: you patched the front door, but forgot to lock the back one.</p><h3>Preventive Measures (OWASP)</h3><p>To avoid this trap:</p><ol><li><strong>Maintain an API inventory</strong>: track all hosts, versions, and environments.</li><li><strong>Decommission old versions</strong>: don’t leave /v1 hanging around.</li><li><strong>Harden non-production APIs</strong>: staging &amp; test should never be public-facing.</li><li><strong>Automate documentation</strong>: use OpenAPI/Swagger with CI/CD pipelines.</li><li><strong>Apply same security controls everywhere</strong>: TLS, auth, logging, and rate limiting.</li><li><strong>Audit regularly</strong>: use scanners (ZAP, Burp, Nuclei, etc.) to detect forgotten endpoints.</li></ol><h3>Summary</h3><p>Improper Inventory Management isn’t flashy. But it’s one of the <strong>easiest mistakes to exploit</strong>. Old APIs, forgotten test environments, and outdated documentation give attackers shortcuts into your system.</p><h3>⚠️ Disclaimer</h3><p>This blog is for <strong>educational purposes only</strong>. The examples shown are meant to help developers and security professionals recognize common mistakes before attackers exploit them.</p><p>If you want to deepen your knowledge, check out <a href="https://www.apisecuniversity.com/"><strong>APIsec University</strong></a> — an amazing resource to learn API security hands-on.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=109a6267049f" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>