Building an Internal Coding Agent at Zup: Lessons and Open Questions

TL;DR

This paper discusses the development of an internal coding agent at Zup, emphasizing engineering choices like tool design and safety measures over prompt engineering to improve reliability and adoption.

Contribution

It highlights the importance of engineering decisions such as tool design and safety guardrails in deploying effective internal coding agents.

Findings

Tool design choices significantly impact agent reliability.

Layered safety guardrails improve trust and safety.

Human oversight modes facilitate organic adoption.

Abstract

Enterprise teams building internal coding agents face a gap between prototype performance and production readiness. The root cause is that technical model quality alone is insufficient -- tool design, safety enforcement, state management, and human trust calibration are equally decisive, yet underreported in the literature. We present CodeGen, an internal coding agent at Zup, and show that targeted tool design (e.g., string-replacement edits over full-file rewrites) and layered safety guardrails improved agent reliability more than prompt engineering, while progressive human oversight modes drove organic adoption without mandating trust. These findings suggest that the engineering decisions surrounding the model -- not the model itself -- determine whether a coding agent delivers real value in practice.