Method

Security Research 2026: Android, Linux, ESP32, OffSec

A living collection of security research notes covering Android internals, Linux kernel hardening, ESP32 firmware analysis, and offensive security techniques.

· updated June 15, 2026

  • #research
  • #android
  • #linux
  • #esp32
  • #offsec

Research areas

This document captures ongoing security research across four domains. Each section links to the relevant tools, PoCs, and writeups.

Android internals

Current focus areas:

  • eBPF-based root detection — how Google’s kernel-level detection works and methods for bypassing it (see the Android Root tutorial)
  • KernelSU internals — kernel patching without modifying system partitions
  • Frida gadget injection — hooking non-debuggable apps via LD_PRELOAD

Key repo: android-reverse-engineering

Linux kernel hardening

Research into CIS benchmarks and kernel parameter hardening:

  • kernel.kptr_restrict=2 — hide kernel pointers
  • kernel.dmesg_restrict=1 — restrict dmesg
  • net.core.bpf_jit_enable=0 — disable BPF JIT
  • kernel.unprivileged_bpf_disabled=1 — restrict BPF to root

Key repo: linux-hardening

ESP32 firmware security

The ESP32 ecosystem has unique attack surfaces:

  • OTA update interception — capturing and modifying over-the-air firmware updates
  • Flash encryption bypass — side-channel attacks on ESP32 flash encryption
  • Bluetooth Low Energy sniffing — passive BLE packet capture with the ESP32’s built-in radio

Key repo: awesome-esp32-security

Offensive security methodology

Tools and techniques in active use:

  • AI-assisted recon — using local LLMs to parse scan data and prioritize targets (see the AI Pentesting Pipeline method)
  • Chain-based reporting — structuring findings as attack chains rather than isolated CVEs (see the Red Team Methodology)
  • Automated PoC generation — template-based exploit scaffolding

Repositories

RepoFocusStars
security-researchCentral research log★1
cve-pocsCVE proof-of-concepts★1
pypentest-aiAI-powered pentesting★0