Agentic RE: Automating Reverse Engineering & Vulnerability Research with AI - John McIntosh - DCTSG2027
Description
Name of Training: Agentic RE: Automating Reverse Engineering & Vulnerability Research with AI
Trainer(s): John McIntosh
Skill Level: Advanced
Dates: March 5 - March 6, 2027
Time: 9:00 am - 5:00 pm
Venue: TBD
Early Bird Cost: S$4,900 (incl. GST, until Dec 31, 2026)
Regular Cost: S$5,200 (incl. GST)
Short Summary
Reverse engineering is entering the Agentic Era. In this two-day, hands-on progression you will pair traditional RE skills with private LLM stacks, custom Model Context Protocol servers, and coding agents extended with your own Agent Skills, and leave with AI-powered workflows that accelerate understanding, surface weaknesses, and streamline your analysis.
Course Description
Reverse engineering is evolving beyond static tools and manual workflows. This two-day, hands-on course introduces a new paradigm: Agentic Workflows. By combining large language models (LLMs), the Model Context Protocol (MCP), and reverse engineering tools like Ghidra, participants learn how to design, configure, and orchestrate AI-powered systems that automate and accelerate binary analysis. Along the way, you turn a coding agent into a reversing agent and capture your RE methodology as reusable Agent Skills.
Rather than focusing on a single operating system, the course demonstrates how agentic pipelines can be applied across Windows, Apple, and Android. You use local models to find real bugs in real targets, including a macOS credential stealer and Windows kernel CVEs from Patch Tuesday, so you learn both how each bug works and how bugs like it are found. Students see how MCP servers and AI-assisted workflows adapt to different platforms, enabling reproducible analysis and vulnerability triage in diverse environments. By the end of the course, participants will have built a reproducible agentic AI workflow capable of analyzing software, surfacing potential vulnerabilities, validating, and triaging results across multiple platforms.
If you are a reverse engineering analyst aiming to speed up your understanding, an AI developer curious about applying agentic workflows to real security problems, or a vulnerability researcher exploring how AI can uncover new weaknesses, this course is built for you.
Course Outline
View full outline
Hide outline
Day 1 - Morning: Foundations and Stack Setup
- The Agentic Era: how LLMs transform reverse engineering and vulnerability research
- LLM basics: tokens, embeddings, quantization, and why models hallucinate CVEs
- Model selection and hardware trade-offs: from Qwen 3.8 27B up to GLM Flash 5.3, DeepSeek 4.1, and the latest local models
- Model Context Protocol (MCP): exposing RE tools to LLMs
- Why local LLMs matter: privacy, reproducibility, control
Day 1 - Afternoon: Foundations and Stack Setup
- Local LLM stack setup: Ollama, OpenWebUI, LM Studio, GhidraMCP, and pyghidra-mcp
- Agentic-assisted reverse engineering: the LLM explains and annotates code
- Real Patch Tuesday CVEs: patch diff a Windows CVE to find the root cause and learn how bugs like it are found
- Using AI to dive into new research areas
- Hands-on lab time and troubleshooting on student stacks
Day 2 - Morning: MCP Development and Agentic Workflow Orchestration
- MCP server basics: Python and FastMCP
- Custom Ghidra MCP with pyghidra-mcp: headless scripting, function listings
- SAST meets LLMs: Semgrep on decompiled code, then LLM triage to find a real kernel integer overflow among the noise
- Lessons from DARPA AIxCC contest winners and insights from agentic bug-finding systems
Day 2 - Afternoon: Coding Agents, Skills, and Orchestration
- Programming with LLMs: DSPy signatures, modules, and ReAct agents that call your Ghidra tools
- Coding agents as reversing agents: let an agent drive the full analysis loop
- Developing Agent Skills: encode your RE methodology as reusable, testable skills
- Securing agentic workflows: prompt injection, sanitization
- RE HUD prototype using Streamlit or Chainlit
- Capstone and proficiency evaluation: students combine deterministic RE tools with agentic reasoning to build reliable, integrated workflows for LLM-enhanced binary analysis, then complete the course CTF and one capstone path
Difficulty Level
Advanced - The student is expected to have significant practical experience with the tools and technologies that the training will focus on.
Suggested Prerequisites
Students should be comfortable with common reverse engineering tools, have a basic understanding of modern AI concepts, and be able to write or modify Python scripts. You do not need deep expertise, but familiarity will help you get the most out of the hands-on work.
- Reverse engineering experience (familiarity with Ghidra, IDA, or similar tools)
- Basic vulnerability research knowledge (common bug classes, analysis workflows)
- Comfort with Python scripting and basic development (installing, debugging, running scripts)
- Experience using recent LLM capabilities in workflows, or at least leveraging AI to improve day-to-day tasks
No pre-work is required. Setup instructions for both the local and the instructor-hosted options are provided before the course, so students arrive with a working environment, and time is reserved on Day 1 for anyone who needs help.
What Students Should Bring
System requirements and alternatives:
- A laptop with 16GB or more of RAM, capable of running Docker and a Linux-style command line
- A GPU is not required. LLM inference is provided by the instructor during the course. Students who want to experiment with local models on their own hardware, for example Qwen 3.8 27B via Ollama, can do so, but this is optional
- Setup instructions for both the local and the instructor-hosted options will be provided before the course
Software requirements:
- Python 3.11 or later
- Docker, to run OpenWebUI and Ollama (non-Docker installation also supported)
- git and a Linux-style command line
What the Trainer Will Provide
Course materials: slides, Jupyter notebooks, annotated scripts, and documentation. Hands-on labs in a preconfigured container with Ghidra, pyghidra-mcp, and every course tool installed, using real-world samples: malware, CVE binaries, and patch diffs. Instructor-hosted LLM inference for the duration of the course, so no student needs a GPU or a paid API account. Optional instructions for connecting the same workflows to frontier model APIs, for students who want to use them. Access to the course CTF server during the class and afterward. No equipment loan or purchase add-on is offered.
Trainer(s) Bio
John McIntosh (@clearbluejar) is a security researcher and lead instructor at ClearSecLabs, a company that offers hands-on training and consulting for reverse engineering and offensive security. He is the author of pyghidra-mcp, a Python-based Ghidra MCP server designed for agentic workflows and reproducible automation, and of ghidriff, a binary patch-diffing engine. Both appear in published frontier-lab n-day exploit research and in independent agentic patch-diffing pipelines. He presented the patch diffing agentic workflow at Objective by the Sea v8 (Reverse Engineering Apple Security Updates).
John has taught related workshops at major conferences, including Supercharging Ghidra: Build Your Own Private Local LLM RE Stack with GhidraMCP, Ollama, and OpenWebUI (Ringzer0 Training), and Offensive Security Tool Development with Ghidra: From Custom CLI Tools to an MCP Server (REcon 2025). With over a decade of offensive security experience, he has spoken and taught at DEF CON Training, Black Hat, Ringzer0, 44CON, Objective by the Sea, and REcon. He regularly blogs on clearbluejar.github.io, publishing detailed write-ups on reversing CVEs and building RE tooling with Ghidra. His teaching emphasizes reproducibility, transparency, and contributor empowerment, bridging deterministic analysis with AI-driven reasoning to accelerate vulnerability research.
Proficiency Exam
Students are evaluated through a two-part hands-on assessment delivered during the course. The first part is a CTF that exercises the full agentic toolchain taught in class, from local LLM stack configuration to MCP-driven binary triage. The second part is an end-to-end agentic workflow capstone, where students choose one of two paths: a Reverse Engineering (RE) path focused on understanding and annotating an unfamiliar binary, or a Vulnerability Research (VR) path focused on surfacing and validating a vulnerability in a target binary. To demonstrate proficiency, students must accumulate a substantial number of points on the course CTF and successfully complete at least one of the two capstone paths (RE or VR). Progress on both is visible to the instructor throughout, so students who fall behind are identified and supported during the course rather than at the end of it.
Registration Terms and Conditions
Trainings are refundable before January 29, 2027, minus a non-refundable processing fee of $350.
Between January 30, 2027 and February 28, 2027 partial refunds will be granted, equal to 50% of the course fee minus a processing fee of $350.
All trainings are non-refundable after Feb 28, 2027.
Training tickets may be transferred to another student. Please email us at training@defcon.org for specifics.
If a training does not reach the minimum registration requirement, it may be cancelled. In the event the training you choose is cancelled, you will be provided the option of receiving a full refund or transferring to another training (subject to availability).
Failure to attend the training without prior written notification will be considered a no-show. No refund will be given.
DEF CON Training may share student contact information, including names and emails, with the course instructor(s) to facilitate sharing of pre-work and course instructions. Instructors are required to safeguard this information and provide appropriate protection so that it is kept private. Instructors may not use student information outside the delivery of this course without the permission of the student.
By purchasing this ticket you agree to abide by the DEF CON Training Code of Conduct and the registration terms and conditions listed above.
Several breaks will be included throughout the day. Please note that food is not included.
All courses come with a certificate, contingent on attendance at 75% of course sessions. Students who pass the end-of-course evaluation will be issued a certificate of proficiency. All other students with sufficient course attendance will be issued a certificate of completion. Students are responsible for making sure their attendance is recorded at the start of the day and after returning from lunch.
