Research Focus · Flagship · Open Work · Research Direction · Contact
I work across security research and software engineering—from reverse engineering and Android runtime analysis to AI agents, multi-agent architectures, and autonomous systems.
- Security R&D — Reverse engineering, Android security, RASP, and code protection.
- AI agents — Multi-agent orchestration, agent architecture, context engineering, and workflow design.
- Autonomous systems — Autonomous VAPT, intelligent threat modeling, specification-driven development, and system-level verification.
An integrity layer for AI-coded software.
Before host agents write, cladding supplies the project's intent. After they work, it checks the result for drift, tests, architecture, and evidence. The goal is simple: turn “done” from a claim into something that can be verified.
Explore cladding · Read the Ironclad standard
- Ironclad — A graded standard for provable consistency among specs, code, and tests.
- harness-boot — A role-focused multi-agent development harness built around living specifications.
- LogcatOn — A high-performance Android logcat viewer for macOS, Windows, and Linux.
Research question: How can autonomous security agents act with enough traceability, containment, and proof to be trusted?
My long-term direction is Autonomous Defense: systems that do more than automate a checklist—they observe, reason, act within boundaries, and leave evidence behind.
“Exploits are written by code. Defenses should be too.”
Open the lab console
$ ./research-loop --status
observe reverse engineering + runtime signals
model threats + system intent
orchestrate focused agents
verify drift + tests + evidence
status building
side_quest cosmic-suika
Off-duty experiment: Cosmic Suika — a 3D space-themed merge game with orbital physics.
For research exchange and open-source collaboration: qwerfunch@gmail.com