CAX-Agent: A Lightweight Agent Harness for Reliable APDL Automation
Quick Answer
CAX-Agent is a lightweight agent harness designed for reliable MAPDL automation, addressing execution control and fault recovery challenges.
Quick Take
CAX-Agent is a lightweight agent harness designed for reliable MAPDL automation, addressing execution control and fault recovery challenges. Empirical evaluation shows that the 'model_only' recovery strategy outperforms others with a completion rate of 92.67% and a task score of 3.59/4 across 450 case runs.
Key Points
- CAX-Agent organizes execution into three layers: LLM service, agent harness, and solver backend.
- The 'model_only' strategy achieved the highest completion rate of 92.67% among three recovery strategies.
- Inter-rater agreement for task completion was strong, with a Cohen's kappa of 0.84.
- The study evaluated 50 standard structural benchmarks with 450 total case runs.
- Findings suggest potential for broader validation of recovery strategies in complex scenarios.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Large language models deployed for MAPDL finite-element simulation face practical reliability challenges: without structured execution control, tool encapsulation, and fault recovery, outputs may be inconsistent and task failures are common. The Agent Harness paradigm addresses this by inserting domain-specific orchestration middleware that manages tool lifecycles, workflow state, and recovery escalation. This paper presents the architecture of CAX-Agent, a lightweight agent harness purpose-built for MAPDL automation, and empirically evaluates one of its core components -- the recovery this http URL-Agent organizes execution into three layers -- LLM service, agent harness, and solver backend -- with a recovery ladder that escalates from deterministic rule patching through model-driven regeneration to context enrichment and human intervention. We evaluate three recovery strategies (no_recovery, rule_only, and model_only) on 50 standard structural benchmarks with three repeated runs per strategy (450 case-runs total). Two independent human raters score task completion under blind conditions; inter-rater agreement is strong (quadratic weighted Cohen's kappa = 0.84, 96 percent of score pairs within one point). Model_only achieves the best completion rate (0.9267), task score (3.59/4), total score (9.16/10), and zero-intervention rate (0.84), outperforming rule_only (0.7733, 3.17/4, 7.03/10, 0.00) and no_recovery (0.6933, 2.74/4, 5.60/10, 0.00) with large effect sizes (Cliff's delta = 0.81-0.87). The benchmark uses deliberately simple geometries to isolate recovery-policy effects; we discuss the scope of these findings and directions for broader validation.
| Comments: | 8 pages, 6 figures, IEEE conference format |
| Subjects: | Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE) |
| Cite as: | arXiv:2605.15218 [cs.AI] |
| (or arXiv:2605.15218v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2605.15218 arXiv-issued DOI via DataCite |
Submission history
From: Yichen Hai [view email]
[v1]
Tue, 12 May 2026 14:46:34 UTC (1,699 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.AI
See more →Procedural Memory Distillation: Online Reflection for Self-Improving Language Models
Procedural Memory Distillation (PMD) enhances reinforcement learning by converting cross-episode signals into reusable memory, improving Qwen3-8B and OLMo3-Instruct-7B models by 3.8-5.5% on SCIKNOWEVAL and 7.9-13.6% on . The co-evolution of policy and memory allows for more effective self-supervision, demonstrating significant performance gains when both components are active.