I work at the intersection of CTF competition systems, AI security, AI-agent engineering, and full-stack product development.
From 2024 to 2025, my work centered on CTF challenge design, competition format design, AI/Web frontier challenge research, next-generation competition platform R&D, and security competition operations at CloverSec. In 2026, I moved into the CloverSec AI Security Lab, focusing on AI security research and applied AI systems.
My current direction is the organic combination of competitions, AI security, and full-stack engineering: using competitions as a way to expose real security problems, using AI agents to improve security workflows, and turning research ideas into systems that can be used by teams under real operational constraints.
- AI Agent + security research: multi-agent CTF solving, coordinator policies, platform polling, challenge import, environment lifecycle management, and writeup generation.
- Competition infrastructure: challenge lifecycle tooling, archive standards, Hub preparation, quality review, final reports, and platform-oriented delivery workflows.
- AI security competitions: AI/Web frontier challenge design, evaluation-oriented competition formats, and tooling that connects research with event operations.
- Full-stack product engineering: backend APIs, frontend interfaces, desktop clients, local storage, packaging, update channels, and release automation for security products.
- Secure delivery systems: Dockerized challenge conversion, platform compatibility checks, reproducible assets, and practical validation pipelines.
- AI application prototypes: internal tools and product experiments that connect AI research with real security-team workflows, including Claw Guard for OpenClaw desktop governance and evidence retention.
- Local-first security tools: native desktop tools for evidence review, WriteUP similarity inspection, fingerprint libraries, runtime checks, and report export.
A Codex plugin marketplace for CloverSec competition work. It connects research, asset collection, resource classification, Dockerizer handoff, writeup generation, archiving, quality review, Hub preparation, and final reporting into one agent workflow.
- GitHub: https://github.com/D1a0y1bb/CloverSec-CTF-ForExample
- Keywords: CTF operations, Codex plugin, MCP, challenge collection, writeup workflow, Docker handoff, Hub preparation.
A Codex skill for turning CTF source and challenge directories into validated Docker deliverables for competition platforms and range engines. It covers common competition scenarios including Web, Pwn, AI, RDG, AWD/AWDP, SecOps, Scenario/Vulhub-like projects, Bundle/Recipe workflows, and Linux-QEMU cases.
- GitHub: https://github.com/D1a0y1bb/CloverSec-CTF-Build-Dockerizer-skill
- Keywords: CTF infrastructure, Docker delivery, challenge validation, competition platform compatibility.
A CTF solving workflow based on ctf-agent, extended for full-event polling, CTFd and Lingxu platform access, manual challenge import, multi-model solving, coordinator policies, environment release, and post-game writeup output.
- GitHub: https://github.com/D1a0y1bb/HuntingBlade
- Keywords: AI-assisted CTF solving, solver swarms, CTFd, Lingxu, multi-model agents, writeup automation.
A native macOS WriteUP audit workbench for comparing text, code, images, metadata, and historical fingerprints, then exporting structured evidence reports for human review.
- GitHub: https://github.com/D1a0y1bb/PitcherPlant
- Keywords: macOS, SwiftUI, WriteUP audit, similarity review, evidence export.
-
Rmitting - a macOS reminders app backed by Apple Reminders, with local organization metadata, archive, recent delete, settings, menu commands, toolbar, and sidebar.
https://github.com/D1a0y1bb/Rmitting -
Codex pet packages - installable Codex pet packages with metadata, transparent animated spritesheets, validation output, contact sheets, prompts, and previews.
https://github.com/D1a0y1bb/Diaoyi-Baobao-Codex-Pet
https://github.com/D1a0y1bb/Tian-Hua-Hua-Codex-Pet
Some of my work is internal to CloverSec, including next-generation competition platform R&D, AI/Web frontier challenge development, event operations, AI Security Lab research prototypes, and Claw Guard, a desktop application for OpenClaw environment identification, health checks, installation, governance, updates, and evidence retention. I only link public repositories here.
Python, Swift, TypeScript, Go, Docker, React, Vite, Wails, SwiftUI, SQLite, macOS, Linux, Windows, Codex plugins, MCP, and agent workflows.
I am interested in AI security, CTF infrastructure, AI-agent systems for security work, full-stack security products, competition platform design, and research-to-engineering projects that can survive real event pressure.