Launching Ikotas Exploit Development Training, Our Practical Exploit Development Program

Overview

We are launching Ikotas Exploit Development Training, a practical training program centered on LLM-assisted exploit development. Covering the major targets researchers work against from an attacker's perspective, including Linux and Windows userland and kernel, browsers, virtualization platforms, and mobile operating systems, it walks through the full flow from vulnerability analysis to constructing exploits that actually succeed, hands-on. The training is built to order, so you can use it for technical development of participants both inside and outside your organization.

A Training Only We Can Deliver

Exploit development has a large wall between the stage of following steps in public write-ups and the stage of building an exploit that stably succeeds against modern mitigations. Reaching arbitrary code execution or privilege escalation from a limited set of primitives under layered defenses such as KASLR, CET, SMAP/SMEP, CFI, and PAC requires target-specific knowledge that is hard to obtain from public information.

We have continuously run, end-to-end and in-house, the full loop from vulnerability discovery through building the working exploit to responsible disclosure, including privilege escalation (0-day/n-day) in the Linux kernel. This training is a structured presentation of the field experience we have built up along the way, together with the practical know-how of each stage: reverse engineering, primitive construction, and stability verification.

Program Structure

Participants choose and combine themes from the list below according to their own interests and work area. The vulnerability classes covered, the exercise environment, and the pace are built to order for each participant to match their current skill level. Each theme is designed so that, by the end of the session, participants can independently construct exploits against similar vulnerabilities in their own environment, covering everything from the fundamentals to breaking through modern mitigations in progressive steps. The tailored training also supports rights transfer, so it can be reused many times as internal or external training rather than being a one-off session. Sample exercises for each theme are listed below.

Linux Userland Exploit

• Fundamentals and applications of ROP (return-oriented programming) and shellcode
• Attacks that exploit the glibc heap allocator
• Bypasses of mitigations such as ASLR/PIE, NX, and CET

Linux Kernel Exploit

• Fundamentals and applications of privilege escalation via Linux kernel vulnerabilities
• Attacks that leverage the internals of memory allocators such as SLUB and sheaf (e.g. cross-cache attacks)
• Attacks against vulnerabilities in Linux kernel-specific facilities such as eBPF and io_uring
• Bypasses of mitigations such as KASLR and SMAP/SMEP

Windows Userland Exploit

• Fundamentals and applications of ROP (return-oriented programming) and shellcode
• Attacks that leverage Windows-specific facilities such as SEH and SegmentHeap
• Bypasses of mitigations such as ASLR, DEP, and CFG

Windows Kernel Exploit

• Fundamentals and applications of privilege escalation on Windows 11
• Attacks against heap vulnerabilities in the pool
• Bypasses of mitigations such as KASLR, SMEP, CFG, HVCI, and VBS

Browser Exploit

• Fundamentals and applications of attacks against JS-engine-specific vulnerabilities such as JIT-engine issues and side effects
• Attacks against V8 Sandbox (Ubercage) vulnerabilities
• Browser process compromise via Mojo IPC and full-chain construction

VM Escape

• Fundamentals and applications of host-side compromise via QEMU/KVM vulnerabilities
• The mechanisms of device emulation (virtio, NVMe, and others) and attacks against them
• n-day development against representative implementations such as VMware and VirtualBox

Android/iOS Exploit

• Fundamentals and applications of rooting via Android kernel (Linux) vulnerabilities
• Attacks against GPU driver vulnerabilities
• Bypasses of mitigations such as SELinux, PAC, and PPL/SPTM

Fuzzing

• Fundamentals and applications of bug discovery with major fuzzers (AFL++, libFuzzer, and others)
• Use of grammar-based and structure-based fuzzers, as well as kernel fuzzers (syzkaller and others)
• Efficient harness writing, identifying causes of coverage stagnation, and crash triage

Our Position

Exploit development is an area where public information alone is difficult to lift to a working-engineer level. LLM-driven vulnerability discovery, too, cannot reach real vulnerabilities without design tailored to the target domain (agent composition, exploration strategy, methods for verifying exploitability). Speaking from a position where we have run the full loop, end-to-end and in-house, from 0-day discovery through building the working exploit to responsible disclosure, we have structured this training to include both knowledge that cannot be obtained from public information alone and the design know-how for putting LLMs to work. The program is aimed at vulnerability researchers, red teamers, and product security engineers who want to raise the defensive posture of their organization or product from an attacker's perspective.

Inquiry and Enrollment

The curriculum is built to order for each participant. We welcome inquiries about enterprise and team enrollment, as well as individual requirements that go beyond the listed themes. Please feel free to contact us.