context-rot-detection

Context Rot Detection is an MCP (Model Context Protocol) server that provides AI agents with real-time self-awareness regarding their cognitive state, particularly focusing on detecting and mitigating performance degradation (context rot) as the context window fills up in long-running sessions. It offers health scoring, model-specific degradation curves, actionable recommendations, session fatigue analysis, and detailed agent health history. It supports integration with numerous LLMs, handles HuggingFace models automatically, and provides both quick and advanced MCP client configurations for seamless integration into agent workflows.