# CODESTRUCT: Code Agents over Structured Action Spaces

arXiv: https://arxiv.org/abs/2604.05407 · HTML: https://mlsanity.com/paper/2604.05407/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Placed in Software engineering coding agents, one of 16 directions in LLM agents and tool use.
- Builds on 11 of its 29 references that are in the map.
- First posted on arXiv 7 Apr 2026.

## Placed in

- LLM agents and tool use / [Software engineering coding agents](https://mlsanity.com/agents/software-engineering-coding-agents/)

## Builds on (most cited first, 11 of 11)

- [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023)
- [SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering](https://mlsanity.com/paper/2405.15793/) (2024)
- [OpenHands: An Open Platform for AI Software Developers as Generalist Agents](https://mlsanity.com/paper/2407.16741/) (2024)
- [Agentless: Demystifying LLM-based Software Engineering Agents](https://mlsanity.com/paper/2407.01489/) (2024)
- [code2vec: Learning Distributed Representations of Code](https://mlsanity.com/paper/1803.09473/) (2018)
- [Graph-based, Self-Supervised Program Repair from Diagnostic Feedback](https://mlsanity.com/paper/2005.10636/) (2020)
- [Abstract Syntax Networks for Code Generation and Semantic Parsing](https://mlsanity.com/paper/1704.07535/) (2017)
- [A Retrieve-and-Edit Framework for Predicting Structured Outputs](https://mlsanity.com/paper/1812.01194/) (2018)
- [Break-It-Fix-It: Unsupervised Learning for Program Repair](https://mlsanity.com/paper/2106.06600/) (2021)
- [CodeAssistBench (CAB): Dataset & Benchmarking for Multi-turn Chat-Based Code Assistance](https://mlsanity.com/paper/2507.10646/) (2025)
- [PSCS: A Path-based Neural Model for Semantic Code Search](https://mlsanity.com/paper/2008.03042/) (2020)

## Built on by (most cited first, 5 of 5)

- [Dissecting model behavior through agent trajectories](https://mlsanity.com/paper/2606.17454/) (2026)
- [CORVUS: Context Optimization and Reduction Via Underlying Synchronization for LLM Coding Agents](https://mlsanity.com/paper/2607.22711/) (2026)
- [QuoteBench: How Matched Scores Can Hide Command-Path Failures](https://mlsanity.com/paper/2608.13547/) (2026)
- [CodeNib: A Multi-View Data System for Serving Repository Context to Coding Agents](https://mlsanity.com/paper/2607.25431/) (2026)
- [Improving Auto-Design of Neural PDE Solvers with a Domain-Specific Language](https://mlsanity.com/paper/2608.04384/) (2026)

## Abstract (arXiv)

LLM-based code agents treat repositories as unstructured text, applying edits through brittle string matching that frequently fails due to formatting drift or ambiguous patterns. We propose reframing the codebase as a structured action space where agents operate on named AST entities rather than text spans. Our framework, CODESTRUCT, provides readCode for retrieving complete syntactic units and editCode for applying syntax-validated transformations to semantic program elements. Evaluated on SWE-Bench Verified across six LLMs, CODESTRUCT improves Pass@1 accuracy by 1.2-5.0% while reducing token consumption by 12-38% for most models. Models that frequently fail to produce valid patches under text-based interfaces benefit most: GPT-5-nano improves by 20.8% as empty-patch failures drop from 46.6% to 7.2%. On CodeAssistBench, we observe consistent accuracy gains (+0.8-4.4%) with cost reductions up to 33%. Our results show that structure-aware interfaces offer a more reliable foundation for code agents.