MCP server · Analytics
Context Rot Detection MCP server
Lets your AI check how healthy its own memory is during long chats and get tips to recover.
This is a small helper that lets your AI check its own mental state during a long conversation. When a chat runs very long, AI assistants quietly get worse at remembering and reasoning, and they don't notice it themselves. This tool gives them a health score and simple advice on what to do about it.
What is an MCP server? The 30-second version
On its own, your AI can only chat with you. An MCP server is a small helper program that gives your AI one new skill or connection. This one connects your AI to a health checker for its own memory, so it can measure how full and how tired its context is. When you ask, the AI calls this helper, gets a score back, and can tell you what it found.
What this MCP server does
You ask your AI to check how it is doing, or it decides to check on its own during a long session. The AI sends this helper a few numbers, like how many tokens are in the conversation and how long the session has been running. The helper compares those numbers to known limits for that model and works out a health score from 0 to 100. It then returns a short report with warnings and simple suggestions, like summarizing older parts of the chat or saving important notes somewhere else. Your AI can show you that report or act on it.
Click to zoomWhat you can do with it
- Check the current health of the conversation and get a 0 to 100 score
- See how full the context window is compared to the model's safe limit
- Get warnings about the lost-in-the-middle effect, where middle details get forgotten
- Get simple recovery tips like compacting context or breaking work into smaller tasks
- Look up past health checks for a specific agent by its ID
- See overall usage statistics across all agents that have used the service
- Use any HuggingFace model name and have the context size looked up automatically
Try asking your AI
- “Check my context health, I think this session is getting long”
- “How full is your context window right now and what is the health score”
- “Show me the health history for agent research-bot”
- “What are the service-wide stats for context rot detection”
What it gives back to you
You get a short report in the chat with a health score from 0 to 100, a status word like healthy or warning, and a breakdown of token use, quality estimates, and session fatigue. It also lists recommendations with a priority and a reason, for example compact context or offload to memory. If you ask for history, you get a list of past checks for that agent. If you ask for service stats, you get totals, averages, and recent activity.
Before you start
What you need
- Node.js 18 or newer if you run it with npx or npm
- Docker if you prefer the container version
- The Claude desktop app or Claude Code, or another MCP-compatible client
Good to know
The health scores are estimates based on research and model profiles, not exact measurements, so treat them as helpful hints rather than hard facts.
Install it with your AI
Add Context Rot Detection MCP server to your AI, no technical skills needed
You don't install anything by hand. You copy one prompt, paste it into an AI that can work on your computer, and it checks, installs and connects the server for you, asking you when it needs something.
Sign in to get the install prompt
Members get a ready-made prompt that lets the Claude desktop app check Context Rot Detection MCP server, install it and connect it for them, step by step. You don't need any technical skills: you copy, paste and answer a few questions. Your connected AI can also find and install any of the 4,066 MCP servers here for you.
Who it's for
People running long AI sessions, developers building agents, and anyone who notices their AI getting forgetful or sloppy after a while.
