Sign InOpen Brain
AI EngineerVideoSource Linked

Building ambitious software — Jonathan Kelley, Dioxus Labs & Cognition

Dioxus found that generating more Rust was easy but merging it was not. Agents paid off on bounded research and maintenance work when architecture, real-device tests, and line-by-line review stayed human-owned.

AI Engineer · Sep 11, 2026
Open Source Open MarkdownOpen JSON
Source Summary

After generating **tens of thousands of lines of Rust**, the Dioxus team found that very little met its merge bar. A Kotlin and Swift plugin implementation arrived quickly, but the team spent **two to three weeks** building test cases and validating it on real devices.

Practical Implication

Use agents where patience and broad technical recall matter: documentation research, debugging specifications, release checklists, backports, and documentation consistency. Keep architecture explicit, communicate intent precisely, and review shipped code line by line.

Agent-Ready Context
After generating **tens of thousands of lines of Rust**, the Dioxus team found that very little met its merge bar. A Kotlin and Swift plugin implementation arrived quickly, but the team spent **two to three weeks** building test cases and validating it on real devices.

Use agents where patience and broad technical recall matter: documentation research, debugging specifications, release checklists, backports, and documentation consistency. Keep architecture explicit, communicate intent precisely, and review shipped code line by line.

Test generation remains a weak point: agents can cover an API while missing the behavior that matters. The experience is from one foundational open-source project, so its workflow may not transfer directly, but its core constraint is clear: **code is cheap; quality is not**.
Connected Context · Feed7 Judgment

This replaces generation volume with mergeability as the meaningful measure of coding-agent value. Against the candidates, it reinforces bounded tasks, explicit architecture, qualified behavioral tests, and reviewable evidence, while resisting any general move away from line-level review for foundational code. The real-device validation cost also shows that fast implementation can merely relocate effort into verification rather than reduce it.

ExecCritic: Learn to Test, Test to Improve for Coding AgentsExecCritic gives experimental support to the reported test weakness: generated tests can harm repair unless test creation is separated, qualified, and frozen before implementation changes.How to Kill the Code Review — Ankit Jain, AviatorAviator proposes shifting review toward intent and verification evidence, whereas Dioxus still found line-by-line review necessary; together they narrow evidence-centered review rather than establishing a universal replacement for diff inspection.How to Get Your Org to Adopt Coding Agents (Without Shipping Garbage) — Eyal Blum, FigmaFigma’s cautious adoption experience reinforces that verification plans, review size, and maintainer feedback—not raw code output—determine whether generated work can be accepted safely.Loop Engineering from First Principles — Kyle Mistele, HumanLayerLoop Engineering turns the Dioxus lesson into an operating constraint: divide work into measured, reviewable increments and pause for human judgment instead of accumulating large volumes of difficult-to-merge code.
Context Map
agentcoding#coding-agents#harness-engineering#agent-reliability
Uncertainty
Test generation remains a weak point: agents can cover an API while missing the behavior that matters. The experience is from one foundational open-source project, so its workflow may not transfer directly, but its core constraint is clear: **code is cheap; quality is not**.