SCOPE: Prompt Evolution for Enhancing Agent Effectiveness

TL;DR

SCOPE introduces an online prompt evolution method for LLM agents, improving their ability to manage dynamic contexts and significantly increasing task success rates without human intervention.

Contribution

It presents a novel online optimization framework with a dual-stream mechanism and perspective-driven exploration for automatic prompt evolution.

Findings

Task success rate increased from 14.23% to 38.64%.

Effective context management reduces errors and enhances agent performance.

Code is publicly available for reproducibility.

Abstract

Large Language Model (LLM) agents are increasingly deployed in environments that generate massive, dynamic contexts. However, a critical bottleneck remains: while agents have access to this context, their static prompts lack the mechanisms to manage it effectively, leading to recurring Corrective and Enhancement failures. To address this capability gap, we introduce \textbf{SCOPE} (Self-evolving Context Optimization via Prompt Evolution). SCOPE frames context management as an \textit{online optimization} problem, synthesizing guidelines from execution traces to automatically evolve the agent's prompt. We propose a Dual-Stream mechanism that balances tactical specificity (resolving immediate errors) with strategic generality (evolving long-term principles). Furthermore, we introduce Perspective-Driven Exploration to maximize strategy coverage, increasing the likelihood that the agent has the correct strategy for any given task. Experiments on the HLE benchmark show that SCOPE improves task success rates from 14.23\% to 38.64\% without human intervention. We make our code publicly available at https://github.com/JarvisPei/SCOPE.