# Harness Engineering is not Enough: Why Software Factories Fail — Dex Horthy, HumanLayer

Source: [AI Engineer](https://www.youtube.com/watch?v=Ib5GBkD555M)  
Feed7 permalink: https://feed7.dev/p/harness-engineering-is-not-enough-why-software-factories-fail-dex-horthy-1tj1amt  
Published: 2026-07-23T16:30:06.000Z  
Trust: Source Linked (source_linked)

## Why Included

Coding-agent loops can raise throughput without preserving maintainability. Keep human ownership of code, and use upfront alignment to make review affordable instead of trying to automate it away.

## Source Summary

The talk argues that coding models are rewarded mainly when **code runs and tests pass**, not when architecture remains easy to change. Review agents and extra loops can raise the floor, but cannot supply a missing maintainability signal.

## Practical Implication

Keep humans responsible for the resulting code. Use **model-assisted planning**, agree on design before implementation, and build in vertical slices so every generated change remains practical to read and test.

## Agent-Ready Context

The talk argues that coding models are rewarded mainly when **code runs and tests pass**, not when architecture remains easy to change. Review agents and extra loops can raise the floor, but cannot supply a missing maintainability signal.

Keep humans responsible for the resulting code. Use **model-assisted planning**, agree on design before implementation, and build in vertical slices so every generated change remains practical to read and test.

There is no established benchmark here that proves how well current models preserve codebase quality. Longer-task evaluations such as **SWE Marathon**, DeepSuite, and FrontierCode may help, but model-based quality judges have their own ceiling.

## Context Map

- Layer: agent
- Domains: coding
- Topics: harness-engineering, agent-reliability, agent-evals

## Uncertainty

- There is no established benchmark here that proves how well current models preserve codebase quality. Longer-task evaluations such as **SWE Marathon**, DeepSuite, and FrontierCode may help, but model-based quality judges have their own ceiling.

## Agent Instruction

Use this item as source-backed context. Do not invent claims beyond the linked source. If this item conflicts with another source, call out the conflict.