EnCompass: Enhancing Agent Programming with Search Over Program Execution Paths
Zhening Li, Armando Solar-Lezama, Yisong Yue, Stephan Zheng
Abstract
We introduce a new approach to *agent programming*, the development of LLM-based agents. Current approaches to agent programming often entangle two aspects of agent design: the core workflow logic and the inference-time strategy (e.g., tree search). We introduce *probabilistic angelic nondeterminism* (PAN), a programming model that disentangles these two concerns, allowing the programmer to describe the agent workflow and independently experiment with different inference-time strategies by simply changing a few inputs. We provide an implementation of PAN in Python as the EnCompass framework, which uses a Python decorator to compile agent workflow programs into a search space. We present three case studies that demonstrate how the framework lets the programmer quickly improve the reliability of an agent and easily switch between different inference-time strategies, all with little additional coding.
BibTeX
@inproceedings{
li2025encompass,
title={EnCompass: Enhancing Agent Programming with Search Over Program Execution Paths},
author={Zhening Li and Armando Solar-Lezama and Yisong Yue and Stephan Zheng},
booktitle={The Thirty-ninth Annual Conference on Neural Information Processing Systems},
year={2025},
url={https://openreview.net/forum?id=IKVkpjSJzJ}
}