I build practical AI agent systems with a focus on software delivery, evaluation, security, and production readiness.
My current work is centered on turning AI-agent ideas into usable engineering workflows: clearer specs, better tool use, measurable evals, safer reviews, and production handoff practices that teams can actually run.
- Agentic software development lifecycle workflows
- AI-agent evaluation, routing, and quality gates
- Security review for tool-using and code-writing agents
- Production readiness checklists and operational playbooks
- Developer experience for Codex, Gemini CLI, MCP, and skill-based agents
Focused Codex/Gemini skills for agentic SDLC work:
agentic-specfor converting ideas into implementation-ready specsagentic-evalsfor evaluation plans, test matrices, and quality gatesagentic-security-reviewfor threat modeling and agent/tool risk reviewagentic-production-readinessfor launch readiness and operational checks
An engineering-first handbook and workspace for AI agent systems, including patterns for context, memory, tooling, evaluation, and production operations.
I care about AI systems that survive contact with real projects. That means fewer demos that only work once, and more attention to:
- precise problem framing
- reproducible validation
- maintainable repo structure
- security and failure modes
- documentation that helps the next builder move faster
I am building public, reusable agent-engineering resources for developers who want to move from prototypes to reliable workflows.
If that overlaps with your work, the best starting point is agentic-sdlc-skills.