OrcaLoca introduces a novel framework that enables LLM agents to accurately pinpoint software issues in complex codebases, addressing a critical gap in Autonomous Software Engineering.

• Integrates LLM capabilities with precise code search mechanisms to improve localization accuracy
• Navigates complex repositories systematically to identify problematic code sections
• Automates issue discovery, reducing developer debugging time
• Enhances software maintenance efficiency through intelligent code exploration

This research represents a significant advancement for engineering teams, enabling more efficient bug detection and resolution in large codebases while reducing manual debugging effort.

OrcaLoca: An LLM Agent Framework for Software Issue Localization