Importance of Human-Readability in Machine-Read Code

Determine how much human-readability matters in research code when an AI agent is the primary reader.

Background

The paper compares code style with reproducibility and maintainability. It notes that, unlike the more direct correspondences between agent-facing repository practices and reproducibility requirements, the role of human-readability is less settled when AI agents increasingly become the primary readers of code. The research-software community has not reached a consensus on the importance of preserving human readability in that setting.

References

Code style maps to readability, a value reproducibility shares but has never placed at its center: it is closer to maintainability than to replication, and the research-software community is itself still debating how much human-readability matters when a machine is the primary reader.

— Reproducibility in the Age of Agentic AI: Context Engineering at the Timescale of a Codebase  (2609.11728 - Barba, 10 Sep 2026) in Section 3, “A Convergence, Not a Coincidence”