@entroly/entroly-mcp

An MCP (Model Context Protocol) server for information-theoretic context optimization. Use it to optimize Claude/Cursor's context window by selecting only the most relevant, high-entropy, and critical code fragments based on a Knapsack constraint.

Installation & Usage

This package is a universal npx bridge to the Entroly Python engine.

You can use it directly in any MCP-compatible client like Cursor or Claude Desktop:

Method 1: entroly-wasm (Recommended — zero dependencies)

{
  "mcpServers": {
    "entroly": {
      "command": "npx",
      "args": ["-y", "entroly-wasm", "serve"],
      "env": {
        "ENTROLY_BUDGET": "200000"
      }
    }
  }
}

Method 2: NPX Bridge to Python (requires pip install entroly)

{
  "mcpServers": {
    "entroly": {
      "command": "npx",
      "args": [
        "-y",
        "@entroly/entroly-mcp",
        "serve"
      ],
      "env": {
        "ENTROLY_BUDGET": "200000"
      }
    }
  }
}

Note: You must have the core Python engine installed on your system:

pip install entroly
# or
pipx install entroly

Features

• Knapsack Token Optimization: Fits the absolute maximum value into your token budget.
• Shannon Entropy Scoring: Prioritizes complex, high-entropy logic over repetitive boilerplate.
• SimHash Deduplication: Never wastes tokens on duplicate file contents.
• Predictive Pre-fetch: Learns your co-access patterns to predict what file you'll need next.
• Feedback Loop: Agentic feedback (record_success / record_failure) continuously tunes the RL weights.

Links

• PyPI: entroly
• Repository: juyterman1000/entroly