# What Do CAE Simulation Agents Really Need Beyond a Generic Harness?

arXiv: https://arxiv.org/abs/2609.03718 · HTML: https://mlsanity.com/paper/2609.03718/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Software engineering coding agents, one of 16 directions in LLM agents and tool use.
- Builds on 18 of its 37 references that are in the map.
- First posted on arXiv 3 Sep 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, 18 of 18)

- [Large Language Model-Based Agents for Software Engineering: A Survey](https://mlsanity.com/paper/2409.02977/) (2024)
- [MechAgents: Large language model multi-agent collaborations can solve mechanics problems, generate new data, and integrate knowledge](https://mlsanity.com/paper/2311.08166/) (2023)
- [MetaOpenFOAM: an LLM-based multi-agent framework for CFD](https://mlsanity.com/paper/2407.21320/) (2024)
- [Dive into Claude Code: The Design Space of Today's and Future AI Agent Systems](https://mlsanity.com/paper/2604.14228/) (2026)
- [Single-Agent LLMs Outperform Multi-Agent Systems on Multi-Hop Reasoning Under Equal Thinking Token Budgets](https://mlsanity.com/paper/2604.02460/) (2026)
- [FEABench: Evaluating Language Models on Multiphysics Reasoning Ability](https://mlsanity.com/paper/2504.06260/) (2025)
- [Rethinking the Value of Multi-Agent Workflow: A Strong Single Agent Baseline](https://mlsanity.com/paper/2601.12307/) (2026)
- [RExBench: Can coding agents autonomously implement AI research extensions?](https://mlsanity.com/paper/2506.22598/) (2025)
- [Foam-Agent 2.0: An End-to-End Composable Multi-Agent Framework for Automating CFD Simulation in OpenFOAM](https://mlsanity.com/paper/2509.18178/) (2025)
- [Integrating Large Language Models for Automated Structural Analysis](https://mlsanity.com/paper/2504.09754/) (2025)
- [ALL-FEM: Agentic Large Language models Fine-tuned for Finite Element Methods](https://mlsanity.com/paper/2603.21011/) (2026)
- [Confucius Code Agent: Scalable Agent Scaffolding for Real-World Codebases](https://mlsanity.com/paper/2512.10398/) (2025)
- [FeaGPT: an End-to-End agentic-AI for Finite Element Analysis](https://mlsanity.com/paper/2510.21993/) (2025)
- [ChatCFD: An LLM-Driven Agent for End-to-End CFD Automation with Structured Knowledge and Reasoning](https://mlsanity.com/paper/2506.02019/) (2025)
- [FEM-Bench: A Structured Scientific Reasoning Benchmark for Evaluating Code-Generating LLMs](https://mlsanity.com/paper/2512.20732/) (2025)
- [OptMetaOpenFOAM: Large Language Model Driven Chain of Thought for Sensitivity Analysis and Parameter Optimization based on CFD](https://mlsanity.com/paper/2503.01273/) (2025)
- [SimuAgent: An LLM-Based Simulink Modeling Assistant Enhanced with Reinforcement Learning](https://mlsanity.com/paper/2601.05187/) (2026)
- [SimBench: A Framework for Evaluating and Diagnosing LLM-Based Digital-Twin Generation for Multi-Physics Simulation](https://mlsanity.com/paper/2408.11987/) (2024)

## Abstract (arXiv)

Computer-aided engineering (CAE) simulation is among the largest and most demanding areas of engineering, where setting up a solver such as OpenFOAM, FEniCS, or COMSOL takes real expertise. Large language model (LLM) agents promise to turn a natural-language request into a working simulation, and recent CAE agents add simulation-specific machinery: multi-agent decomposition, domain retrieval, and scripted reflection. That machinery suited weak base models; modern harnesses already supply multi-turn reasoning, tool use, and execution feedback. We ask what a CAE simulation agent still needs beyond a generic harness. With information access and repair budget held fixed, a single-agent harness matches or beats multi-agent specialized systems (FoamBench 96.4% vs. 88.2%). Ablations trace this to capabilities the harness already provides: execution-feedback repair lifts FoamBench from 71.8% with no repair round to 96.4%, while scripted reflection adds nothing. The one input that still helps is domain knowledge supplied as solver tutorials, our largest measured gain (80.9% to 96.4%).