From Fleet to Flag: An LLM Agent that Autonomously Solves a CTF

I spent a week trying to get a machine to solve a Hack The Box web challenge on its own — no human feeding it the vulnerability, no peeking at the writeup, just source code, a live target, and a model (sometimes handed a lead by another model, never by me). This post is the honest account of what worked, what didn't, and what the numbers actually say about model reasoning, hypothesis seeding, and the role of the harness around the model.

This is not a challenge writeup

The CTF is only the measuring stick. What I actually care about here is the models and the harness: how architecture around a model decides whether it can reason at all, where stronger models pull ahead, how one model's hypothesis accelerates another, and what it costs. The vulnerability details appear only as evidence for those claims — if you want the clean exploit, the payload is in one code block and you can leave.

The headline: a mid-tier open model (DeepSeek v4-pro), handed another model's hypothesis and a thin single-agent loop, autonomously converted it into a working Mongoose prototype-pollution exploit and pulled a live flag for $1.81 — the exact payload every other configuration circled but missed.

Espionage Intelligence Writeup

Summary

Espionage Intelligence was an HTB CTF challenge centered around abusing an Operator Wiki RAG pipeline that ranked documents by semantic similarity while failing to properly enforce clearance boundaries. Starting with a standard HUMINT operator account, the objective was to perform reconnaissance of the semantic space, extract sensitive information useful for lateral movement, escalate privileges, and ultimately access advanced leader-only agentic analytics.

HTB ARMAXIS Writeup

Summary

The application was vulnerable to a logical flaw in the password reset mechanism that allowed unauthorized password resets. By exploiting this flaw, we reset the admin user's password. This privilege escalation allowed access to an admin-only weapon dispatch feature, which included a command injection vulnerability via unsanitized curl execution in Markdown parsing. Combining both vulnerabilities led to command execution and flag retrieval.

  • in Red Team
  • 10 min read

Abusing GitHub Codespaces as a Cloud Based Exfiltration Proxy

INTRODUCTION

GitHub Codespaces is a cloud-based development environment that allows developers to instantly spin up a fully configured, containerized VS Code environment directly from a GitHub repository. It eliminates the need for manual setup and ensures consistency across development environments by using devcontainers. With support for Visual Studio Code and browser-based editing, Codespaces enables fast onboarding, seamless collaboration, and streamlined CI/CD workflows, making it ideal for modern DevOps and remote development scenarios.

Pwning PaaS Cloud Managed Services From Breaking Isolation to Identifying Abuse

Introduction

This research is heavily inspired by the insightful work published by the team at Wiz (https://www.wiz.io/blog/the-cloud-has-an-isolation-problem-postgresql-vulnerabilities). Motivated by their findings, I conducted an independent assessment focused on Platform-as-a-Service (PaaS) environments—specifically targeting platforms that offer managed services in two primary categories:

  • Coding Platforms
  • Managed Service Platforms