Software factory (AI agents)
A pipeline where agents write and fix most of the code while people write specs, set guardrails, design the checks and decide what ships.
The idea takes spec-driven development and loops such as the Ralph loop and runs them as a process: specs go in, background agents implement, verifiers (tests, linters, audits) reject bad work, and a person reviews what passes. The bottleneck moves from typing code to writing good specs and checks. Without strong verifiers it becomes a fast way to produce technical debt.
Related terms
- Spec-driven development: Writing the specification and acceptance tests first, then letting agents implement against them.
- Ralph loop (Ralph Wiggum technique): Running the same prompt against a coding agent again and again, each time with a fresh context, while progress lives in files and tests decide when it is done.
- Background coding agent: A coding agent that runs in the cloud on its own, takes an issue or a task and comes back later with a pull request for review.
- Audit-driven development: A loop for code you maintain: an AI agent audits it against a checklist, the findings become small fixes, and it audits again until nothing is left.
- Self-driving product: A product whose usage data drives an agent loop: find what to fix, open a pull request, measure the change, while engineers set limits and approve merges.
