<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>shulkwisec — Security Researcher &amp; Software Developer</title><description>Security researcher and software developer specialised in exploit development and AI-driven security tooling. Depth across web and binary exploitation, reverse engineering, cryptography, and source-code review.</description><link>https://shulkwisec.github.io/</link><language>en-us</language><item><title>PHP to Shell, Part 1: Building an Isolated Lab Worth Trusting</title><link>https://shulkwisec.github.io/php-to-shell/01-building-the-lab/</link><guid isPermaLink="true">https://shulkwisec.github.io/php-to-shell/01-building-the-lab/</guid><description>A loopback-only PHP target built from native packages, not Docker, with the egress claim proven by seven checks instead of assumed from the config.</description><pubDate>Sat, 14 Nov 2026 00:00:00 GMT</pubDate><category>php-to-shell</category><category>web</category><category>lab-setup</category></item><item><title>PHP to Shell, Part 2: Unrestricted Upload to Webshell</title><link>https://shulkwisec.github.io/php-to-shell/02-upload-to-webshell/</link><guid isPermaLink="true">https://shulkwisec.github.io/php-to-shell/02-upload-to-webshell/</guid><description>An upload bypass needs two independent things to line up — a servable location and a server that routes it to PHP. Tested against a real Apache + mod_php target, including the negative controls that prove which config choice is actually load-bearing.</description><pubDate>Sat, 14 Nov 2026 00:00:00 GMT</pubDate><category>php-to-shell</category><category>web</category><category>file-upload</category></item><item><title>PHP to Shell, Part 3: File Inclusion to Code Execution</title><link>https://shulkwisec.github.io/php-to-shell/03-file-inclusion-to-code-execution/</link><guid isPermaLink="true">https://shulkwisec.github.io/php-to-shell/03-file-inclusion-to-code-execution/</guid><description>An include() on user input is a read primitive until something makes the included content attacker-controlled PHP. Logs, sessions, phar archives and a filter-chain trick that I could only confirm halfway.</description><pubDate>Sat, 14 Nov 2026 00:00:00 GMT</pubDate><category>php-to-shell</category><category>web</category><category>lfi</category></item><item><title>PHP to Shell, Part 4: Object Injection to Code Execution</title><link>https://shulkwisec.github.io/php-to-shell/04-object-injection-to-code-execution/</link><guid isPermaLink="true">https://shulkwisec.github.io/php-to-shell/04-object-injection-to-code-execution/</guid><description>unserialize() on attacker input is a necessary condition, not a finding. A usable gadget chain needs the exact class, the exact property names, and the exact method flow to still be present — and most of the time, they aren&apos;t.</description><pubDate>Sat, 14 Nov 2026 00:00:00 GMT</pubDate><category>php-to-shell</category><category>web</category><category>deserialization</category></item><item><title>PHP to Shell, Part 5: The CVEs That Carry This Chain in the Wild</title><link>https://shulkwisec.github.io/php-to-shell/05-the-cves-that-carry-this-chain/</link><guid isPermaLink="true">https://shulkwisec.github.io/php-to-shell/05-the-cves-that-carry-this-chain/</guid><description>CVSS predicts almost nothing about which of these actually gets exploited. Install base, remote fingerprintability, and whether the upload path runs before the framework&apos;s own protections predict it far better.</description><pubDate>Sat, 14 Nov 2026 00:00:00 GMT</pubDate><category>php-to-shell</category><category>web</category><category>cve</category></item><item><title>PHP to Shell, Part 6: Webshell Payloads, and How They Get Caught</title><link>https://shulkwisec.github.io/php-to-shell/06-webshell-payloads-and-how-they-get-caught/</link><guid isPermaLink="true">https://shulkwisec.github.io/php-to-shell/06-webshell-payloads-and-how-they-get-caught/</guid><description>A webshell is one function call, deliberately. The closing post in the series: the minimal execution primitives, why disable_functions isn&apos;t a boundary, and the detection signal that repeats across every post in this series rather than being specific to any one of them.</description><pubDate>Sat, 14 Nov 2026 00:00:00 GMT</pubDate><category>php-to-shell</category><category>web</category><category>detection-engineering</category></item><item><title>bb-huge: building a senior bug bounty hunter out of six layers</title><link>https://shulkwisec.github.io/03-Projects/use-ai-agents-to-build-autonomously-senior-bug-bounty-hunter-2026/</link><guid isPermaLink="true">https://shulkwisec.github.io/03-Projects/use-ai-agents-to-build-autonomously-senior-bug-bounty-hunter-2026/</guid><description>A senior hunter is mostly carried context: the methodology a session starts with, the memory that outlives it, the tools it can call, and the clock that runs unattended. Six independent layers.</description><pubDate>Tue, 09 Jun 2026 00:00:00 GMT</pubDate><category>bb-huge</category><category>ai</category><category>agent</category><category>bug-bounty</category><category>ai-for-sec</category><category>orchestration</category><category>sub-agents</category><category>autonomous-ai</category></item><item><title>emmu: running an LLM as an operator instead of a chat partner</title><link>https://shulkwisec.github.io/03-Projects/emmu-the-autonomous-cyber-ai-bridge/</link><guid isPermaLink="true">https://shulkwisec.github.io/03-Projects/emmu-the-autonomous-cyber-ai-bridge/</guid><description>emmu turns a frontier model into an operator: 60+ specialist agents, a reviewer gate in front of every tool call, and a workspace that survives a crash. How it is put together, and where a human still has to be in the loop.</description><pubDate>Fri, 08 May 2026 00:00:00 GMT</pubDate><category>emmu</category><category>ai-agent-framework</category><category>ai-orchestration</category><category>mcp-bridge</category><category>autonomous-hacking</category><category>ctf-solver</category></item><item><title>OSINT Framework: turning a username into a footprint</title><link>https://shulkwisec.github.io/02-Research/osint-framework-by-shulkwisec/</link><guid isPermaLink="true">https://shulkwisec.github.io/02-Research/osint-framework-by-shulkwisec/</guid><description>One Discord handle, one branch of osintframework.com, and a tool called WhatsMyName: how a single string turns into a footprint, and where the result stops being evidence.</description><pubDate>Sun, 25 Jan 2026 00:00:00 GMT</pubDate><category>osint</category><category>framework</category></item><item><title>mal_sim (JCOE_CTF) writeup</title><link>https://shulkwisec.github.io/04-CTF-Writeups/JCOE_CTF_2025/mal-sim-jcoe-ctf-writeup/</link><guid isPermaLink="true">https://shulkwisec.github.io/04-CTF-Writeups/JCOE_CTF_2025/mal-sim-jcoe-ctf-writeup/</guid><description>A puzzle-locked binary that guards a C2 hostname: reading the prompts with strings, solving three locks by hand, and getting the flag out in one run.</description><pubDate>Wed, 29 Oct 2025 00:00:00 GMT</pubDate><category>jcoectf</category><category>ctf</category><category>red-team</category></item><item><title>KashiCTF 2025: carving a flag out of a corrupted VirtualBox disk</title><link>https://shulkwisec.github.io/04-CTF-Writeups/KashiCTF_2025/kashictf-self-destruct/</link><guid isPermaLink="true">https://shulkwisec.github.io/04-CTF-Writeups/KashiCTF_2025/kashictf-self-destruct/</guid><description>VirtualBox refused to open the challenge disk, so the disk was not the puzzle. The payload was still inside the container, and the way in was to stop trusting the metadata and carve the filesystem out by hand.</description><pubDate>Wed, 29 Oct 2025 00:00:00 GMT</pubDate><category>kashictf</category><category>ctf</category><category>exploitation</category><category>red-team</category></item><item><title>KnightCTF 2025: cracking RSA with a predictable prime</title><link>https://shulkwisec.github.io/04-CTF-Writeups/KnightCTF_2025/knightctf-2025-writeups-randomsenioradleman/</link><guid isPermaLink="true">https://shulkwisec.github.io/04-CTF-Writeups/KnightCTF_2025/knightctf-2025-writeups-randomsenioradleman/</guid><description>RSA stays secure as long as the prime is unpredictable. This one generated its prime with a seeded PRNG, and the seed was hiding in plain sight in the challenge hints.</description><pubDate>Wed, 29 Oct 2025 00:00:00 GMT</pubDate><category>knightctf</category><category>ctf</category><category>red-team</category></item><item><title>bugbountysa CLI Tool</title><link>https://shulkwisec.github.io/05-Resources/bugbountysa-cli-tool-post/</link><guid isPermaLink="true">https://shulkwisec.github.io/05-Resources/bugbountysa-cli-tool-post/</guid><description>Turn BugBounty.SA programs into JSON on disk: scopes, domains and bounty ranges you can grep instead of scroll.</description><pubDate>Tue, 28 Oct 2025 00:00:00 GMT</pubDate><category>bugbounty-sa</category><category>cli</category><category>opensource</category><category>security-tools</category><category>automation</category></item><item><title>Getting started with Sahb: your own research site from a folder of Markdown</title><link>https://shulkwisec.github.io/01-Articles/getting-started-useing-sahb-template-by-shulkwisec/</link><guid isPermaLink="true">https://shulkwisec.github.io/01-Articles/getting-started-useing-sahb-template-by-shulkwisec/</guid><description>Sahb is a static research site that publishes itself from a folder of Markdown notes. Clone it, point it at your vault, write a note, and the site builds it.</description><pubDate>Sat, 01 Feb 2025 00:00:00 GMT</pubDate><category>getting-started</category><category>guide</category></item></channel></rss>