Concept

Human role

You might say: owner, product decider, factory engineer, head chef

The work left for people once agents write and prove most changes. People decide what should exist and write the specs. They own the one-way doors and the gates, read the diffs of high-risk changes, watch outcomes and cost, and fix the factory when the same mistake repeats. A person attends to the factory when it stops or asks, the way a Toyota worker attends a machine only when it stops on an abnormality. In kitchen terms the person is the head chef, who designs the recipes while agents cook.

What it does
Puts scarce human attention on the decisions that only a person can make.
Ask
Which decisions here can only a person make, and where is someone still spending attention on work the factory could prove?

Pattern

Stop the line

Connections

Sources

  1. Kent C. Dodds, You Own the Outcome: Product Engineering in the Age of AI, Decide what should existThe valuable part is deciding whether the request should be built at all, and what it should actually be.
  2. Kent C. Dodds, kcd-skills, autopilot skillAgents ship most changes; the human owns product decisions.
  3. Kent C. Dodds, kcd-skills, autopilot skillProduct calls and one-way doors go to the human.
  4. Kody, kody-hq/kody repository, docs/contributing/harness-engineering.mdOptimize for human attention as the scarce resource.
  5. Matt Pocock, AI Coding Dictionary, Software factoryDeciding which of those decisions stay human is the main design question.
  6. Delinea, The Delinea Software Factory: Building Software with AI, The Delinea Software FactoryHumans define intent: what the system should do, the scenarios it needs to handle, the constraints that matter.
  7. Taiichi Ohno, Toyota Production System: Beyond Large-Scale Production, Chapter 1, Starting from Need (Taylor & Francis preview PDF)Only when the machine stops because of an abnormal situation does it get human attention. As a result, one worker can attend several machines, making it possible to reduce the number of operators and increase production efficiency
  8. Kent C. Dodds, You Own the Outcome: Product Engineering in the Age of AI, Watch the outcomesBuild self-healing automations. Things will break. Set things up so that failures get noticed and triaged (often by another agent) instead of silently rotting.
  9. Gene Kim and Steve Yegge, Vibe Coding: Building Production-Grade Software With GenAI, Chat, Agents, and Beyond, From Line Cook to Head Chef: Orchestrating AI Teams (book excerpt)Welcome back, head chefs. You’ve mastered working with your AI sous chef.
  10. Mary Poppendieck and Tom Poppendieck, Lean Software Development: An Agile Toolkit, Introduction, p. xxvi (publisher's sample chapter)Development is like creating a recipe, while production is like making the dish.

Use it on your project

teardown-apply: hold "Human role" (human-role) from the teardown at https://teardownatlas.com/software-factory/ against my project

Paste it into a coding agent that has the teardown skills.

An AI agent wrote this map with the teardown skill. It can contain errors: check the sources before you rely on a part. About and legal