← All tracks
Shipping with AI coding agents
Go from pasting prompts into a chat window to running a repo where agents do real, reviewable work.
Scope
What we work on
- Agent edits that look right, pass review, and quietly break production
- No repeatable workflow — every task starts from a blank prompt
- Context blowouts: huge token bills, worse answers
- Multi-agent setups that burn money without shipping anything
- No idea which work to hand to an agent and which to keep
What you walk away with
- A repo configured for agents: instructions, skills, hooks, guardrails
- A personal workflow for planning → building → reviewing → landing changes
- Review habits that catch the specific ways generated code fails
- A cost model you can defend to whoever pays the bill
Who this is for
- Engineers who use Claude Code or Cursor daily but feel like output plateaued
- Tech leads deciding how their team should adopt agentic tooling
- Solo builders trying to ship more surface area than one person should be able to
Why me for this
I run a multi-repo platform where the majority of code is written by agents under a review pipeline I built. The failure modes I teach are ones I paid for first.
Next step
Start with a free 20-minute intro call
No pitch. We work out whether I can actually help, and I'll say so if I can't.