I build security products, public-safe release hubs, and practical Codex workflows that help good work ship clearly.
EN
TR
DE
ES
PT-BR
FRMost of the work is organized around a few connected projects. If you are visiting for the first time, this is the quickest way to find the right one:
| If you need... | Start with |
|---|---|
| Vulnerability intelligence, CVE workflows, and public product docs | MyVuln live product and MyVuln public hub |
| Public releases, signed update manifests, and operator support evidence | VaultPilot release hub |
| Windows-first Codex setup with agents, skills, MCP connectors, and validation gates | Codex Chef |
| Reusable approval-gated prompts, workflows, and verification checklists | Prompt Architect |
| Documentation-first starter kits for AI-coded projects | AI Project Starter |
| Reusable AI agent and Codex skill creation | AI Skill Create |
| Codex CLI operating guidance for prompts, skills, MCP, hooks, and Windows | Codex CLI Operator Handbook |
| Turkish cybersecurity writing and public research surface | SiberDergi |
- Security products and public hubs that make useful documentation, releases, and trust signals easy to find without exposing private source, customer data, or tenant context.
- Windows-first release surfaces for self-hosted tools, with the manifests, support evidence, operator notes, and verification gates needed to keep shipping understandable.
- Codex and AI-engineering tools that turn a vague idea into a clear plan, a reusable skill, a well-routed workflow, and a handoff another person can actually pick up.
    |     |
    |     |
Alongside the products, I build the small systems that make AI-assisted engineering more deliberate: better project context, safer prompts, reusable skills, and workflows that leave evidence behind.
    |     |
    |     |
| Loop | Proof in the repo surface |
|---|---|
| Research | I read the public docs, threat boundaries, release notes, and repo instructions before changing behavior. |
| Plan | I keep the scope clear, define the checks up front, and leave room for a safe rollback. |
| Execute | I make the focused change without weakening auth, validation, public-safety, or support boundaries. |
| Verify | I start with local checks, then widen the evidence when the change can affect more of the system. |
| Release | I keep the docs clean and make the public artifact easy for the next person to trust and use. |
| Area | Tools and platforms |
|---|---|
| Product UI and docs | TypeScript, React, Next.js, Tailwind CSS, Markdown, SVG assets |
| Security products | CVE workflows, URL intelligence, release manifests, support evidence, public-safe docs |
| Data and backend | Node.js, Supabase, PostgreSQL, SQLite, Prisma |
| Windows release work | PowerShell, Windows Server, signed release flow, GitHub Actions |
| AI engineering | OpenAI Codex, skills, agents, MCP connectors, prompt architecture, verification checklists |
The contribution snake is kept as a manual-only profile asset. There is no cron trigger and no push trigger quietly changing the profile in the background.
This profile points to public-safe hubs on purpose. It is not the place for private source paths, customer data, tenant screenshots, incident details, tokens, credentials, local machine paths, or private operational context.
Build. Verify. Release.
