ReCreate: Reasoning and Creating Domain Agents Driven by Experience

2026-01-16

AI Summary

This work proposes an experience-driven framework for autonomous agent self-optimization, addressing the limitations of existing approaches that either rely heavily on manual design or suffer from high computational costs and a lack of causal attribution in success or failure. The framework adopts an “agent-as-optimizer” paradigm, leveraging interaction histories to automatically distill reusable domain-specific patterns. It integrates an experience memory with retrieval mechanisms, a reasoning-generation co-pipeline, and a hierarchical abstraction update strategy, enabling continuous improvement starting from a minimal initial template. Experimental results demonstrate that the method significantly outperforms both handcrafted agents and current automated agent construction techniques across multiple tasks, efficiently producing high-performance agents with minimal human intervention.

Technology Category

• Search and Optimization: Learning to Search
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Multiagent Systems: Mechanism Design

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Large Language Model agents are reshaping the industrial landscape. However, most practical agents remain human-designed because tasks differ widely, making them labor-intensive to build. This situation poses a central question: can we automatically create and adapt domain agents in the wild? While several recent approaches have sought to automate agent creation, they typically treat agent generation as a black-box procedure and rely solely on final performance metrics to guide the process. Such strategies overlook critical evidence explaining why an agent succeeds or fails, and often require high computational costs. To address these limitations, we propose ReCreate, an experience-driven framework for the automatic creation of domain agents. ReCreate systematically leverages agent interaction histories, which provide rich concrete signals on both the causes of success or failure and the avenues for improvement. Specifically, we introduce an agent-as-optimizer paradigm that effectively learns from experience via three key components: (i) an experience storage and retrieval mechanism for on-demand inspection; (ii) a reasoning-creating synergy pipeline that maps execution experience into scaffold edits; and (iii) hierarchical updates that abstract instance-level details into reusable domain patterns. In experiments across diverse domains, ReCreate consistently outperforms human-designed agents and existing automated agent generation methods, even when starting from minimal seed scaffolds.

Problem

• agent creation
• domain adaptation
• experience-driven
• automated reasoning
• large language models

Innovation

• experience-driven agent creation
• reasoning-creating synergy
• agent-as-optimizer
• hierarchical abstraction
• interaction history