SCOPE: Prompt Evolution for Enhancing Agent Effectiveness

2025-12-17

AI Summary

To address the failure of static prompting in large language model (LLM) agents operating within dynamic, massive-context environments—leading to ineffective error correction and capability enhancement—this paper proposes SCOPE, an execution-trajectory-driven online prompt evolution framework. Methodologically, SCOPE integrates online optimization, execution trace analysis, neuralized prompt tuning, and multi-perspective strategy sampling. Its key contributions are: (1) a novel dual-stream prompt evolution mechanism that jointly optimizes tactical-level real-time error correction and strategic-level cross-task generalization; and (2) a perspective-driven exploration strategy enabling autonomous, human-intervention-free context management. Evaluated on the HLE benchmark, SCOPE significantly improves task success rate from 14.23% to 38.64%. The implementation is publicly available.

Technology Category

• Natural Language Processing: Prompt Engineering / Prompting
• Search and Optimization: Learning to Search
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

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 extbf{SCOPE} (Self-evolving Context Optimization via Prompt Evolution). SCOPE frames context management as an extit{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.

Problem

• LLM agents lack dynamic prompt mechanisms for massive contexts
• Static prompts cause recurring corrective and enhancement failures
• Need automated prompt evolution to optimize context management

Innovation

• Self-evolving prompts via online context optimization
• Dual-Stream mechanism balancing tactical and strategic updates
• Perspective-Driven Exploration to maximize strategy coverage