Unity Insight improves code asset search for game developers
Unity Insight: A Production Code--Asset Index for LLM Coding Agents in Unity Projects
Software Engineering
Summary
Game development in Unity involves many connected files that store code and game data. The authors found that existing tools struggle to search across these linked code and asset files because relationships are hidden in non-code files. They created Unity Insight, a tool that makes it easier for AI agents to quickly find connections across all Unity files. When tested on two games, Unity Insight helped AI agents answer questions using fewer tokens and less time than general methods.
What this means in practice
- •For game developers: Find dependencies and connections across code and assets quickly in complex Unity projects using AI tools powered by Unity Insight.
- •For software engineering teams: Reduce time and compute resources needed for automated code understanding and refactoring in projects mixing code and serialized data.
Authors
Shenhua Gu, Hongqiang Zhu, Fan Zhang, Jinming Zhang, Hao Chen
Abstract
LLM coding agents increasingly operate inside game-engine repositories, where application logic is inseparable from serialized assets: a single gameplay change may span C\# scripts, prefabs, scenes, and ScriptableObjects wired together by Unity GUIDs. The retrieval tools agents carry today---shell utilities and code-only indexes---cannot answer basic cross-file questions, because these relationships live in \texttt{.meta} files and YAML assets rather than in code. We present Unity Insight, to our knowledge the first persistent, LLM-facing, agent-integrated cross-file code--asset index for Unity projects, shipping in production with Tuanjie Codely, the agent CLI of Tuanjie Engine, since its public launch on 2026-07-28. In a paired experiment---28 project-specific questions on two Unity games, same model and harness, one run per arm per question---the index-backed agent spent 53\% fewer tokens and 52\% less wall-clock time than a general-purpose exploration agent (exact paired sign tests, $p{<}0.004$), using only its typed index-query tools.