Daniel Sogl

Telling your AI coding agent "please don't commit directly to main" is about as effective as asking your cat not to knock things off the table. LLMs are non-deterministic by nature - a polite prompt instruction is not a security boundary.
In this session, you'll learn how to enforce what you can't simply ask for. We'll dive into lifecycle hooks that intercept tool calls before they execute, blocking unauthorized Git commits, file deletions, or rogue shell commands at the architectural level. We'll also explore how to run coding agents autonomously and safely inside sandboxes - so your agent can do real work without being a liability.

Whether you're building agentic workflows or just trying to stop your AI colleague from pushing directly to production: this talk gives you the controls that prompts never will.

AI-generated code is becoming the norm, but who reviews the reviewer? In this session, we explore how to close the feedback loop by letting AI agents review AI-written code. We'll look at local agent setups as well as cloud-based services like GitHub Copilot code review or Greptile, and discuss when each approach makes sense. Walk away with a practical mental model for building a self-correcting AI development workflow, without losing control over your codebase.

Developers now create entire applications by simply chatting with AI tools. More and more startups rely on predominantly AI-generated code. But what initially thrilled developers quickly leads to frustration: generic code, poor architecture, and security vulnerabilities.In this talk, I demonstrate how Spec Driven Development (SDD) solves these problems. Instead of coding first and documenting later, SDD begins with precise specifications as the "single source of truth." Various tools and CLIs make this methodology practically applicable today with a structured 4-phase approach: Specify, Plan, Tasks, Implement.Using a real project, I show how SDD works with modern AI coding tools. I share honest insights from hands-on experience: where SDD excels, where it fails, and whether the specification overhead actually leads to better results.