Cole Medin

Coding agents like Claude Code barely write bad code anymore. They write good code that misses the point. The failure moved from code time, where the agent couldn't write the function, to intent time, where it wrote a perfectly fine function for a task that wasn't quite the one in your head. That's misalignment, not incompetence.

Our reflex is more review, more tests, more guardrails. All of it runs downstream of the failure. The miss already happened before any of those fire, so you're doing an autopsy instead of preventing anything.

This talk shows the fix: work backwards. Before your agent writes a single line, you and the agent together define concrete, measurable success criteria and the validation strategy that proves them. Only then does it implement. It's TDD repurposed for AI agents! The validation-first plan gives the AI agent an actual contract for the task at hand.

I'll walk through it live on a real codebase, in Claude Code, on an ordinary task: the naive version that passes every check and still does the wrong thing, the plan that makes "done" explicit, and the same task done validation-first. Then the part that compounds, which is that when a miss does slip through you don't just patch the bug, you evolve the system so the same class of mistakes goes away over time. You own the plan, the agent owns the typing.

Attendees leave with three things they can apply on their next ticket, with no new tooling required.

I have been running an AI second brain as a developer for the past year, and it saves me at least 20 hours every week. It knows my repos and the conventions in each one, dispatches coding work to run in the background, drafts my email replies, and pings me on Slack when something needs my attention. In this talk I will show you how to build one in TypeScript.

We will walk through the four parts that make it work. The memory is just markdown files that the agent writes to and curates itself. Hybrid search runs over that memory so the agent pulls the right context at the right time. A scheduled loop makes it proactive, reviewing what happened yesterday, promoting what mattered into long term memory, and surfacing what needs me today. And an integration layer connects all of it to my repositories, my inbox, and the rest of the platforms I actually work in.

You will leave with the full architecture and a clear path to building your own.

You'll learn how to move beyond basic prompting to create comprehensive workflows to give your AI coding assistants all the architecture, best practices, project rules, code examples, and validation processes it needs to get the job done effectively. Through hands-on exercises with Claude Code, we'll demonstrate how proper Context Engineering unlocks a whole new level of agentic engineering while teaching practical techniques, templates, and best practices that help AI assistants understand your project standards and deliver production-ready code consistently.