RuntimeErrorSage: Intelligent Runtime Error Analysis and Remediation using Local Large Language Models

RuntimeErrorSage is a .NET middleware system designed to intelligently analyze and remediate runtime errors using a locally hosted Large Language Model (LLM), specifically Qwen 2.5 7B Instruct 1M. Unlike cloud-based tools, RuntimeErrorSage operates entirely offline, ensuring data privacy and low latency error resolution. It intercepts unhandled exceptions, generates structured context via the Model Context Protocol (MCP), and utilizes an LLM to suggest and execute safe remediation actions. This research presents a formal system model for runtime error classification, context graph analysis, and remediation decision-making. Evaluation results show 92% classification accuracy and 85% remediation success with an average resolution time of 2.3 seconds. The system is open source and designed for integration into ASP.NET Core applications, offering production-grade reliability without external dependencies.