Skip to main content
Glama
ZeroEntropy Zerank MCP Server

ZeroEntropy Zerank MCP Server

by jbccc

ZeroEntropy Zerank MCP Server

A Model Context Protocol (MCP) server for document reranking using the ZeroEntropy Zerank API, now implemented in Node.js.

Features

  • Document reranking using the ZeroEntropy Zerank API

  • Input validation with Zod schemas

  • Support for both standalone Node.js server and Cloudflare Workers deployment

  • MCP protocol compliance for integration with AI assistants

Related MCP server: Cohere MCP Server

Installation

npm install

Usage

Standalone Node.js Server

Run the MCP server using stdio transport:

npm start

Or for development with auto-reload:

npm run dev

Cloudflare Workers Deployment

  1. Install Wrangler CLI:

npm install -g wrangler
  1. Login to Cloudflare:

wrangler login
  1. Deploy the worker:

wrangler deploy

API

The server provides one tool:

get_reranking

Reranks a list of documents based on relevance to a query.

Parameters:

  • query (string): The search query (1-10000 characters)

  • documents (array): Array of documents to rerank (1-1000 items)

  • api_key (string): ZeroEntropy API key

Returns:

  • results (array): Array of reranked results with index and relevance_score

Example:

{
  "query": "machine learning algorithms",
  "documents": [
    "Neural networks are a type of machine learning model",
    "Cooking recipes for pasta dishes",
    "Supervised learning techniques in AI"
  ],
  "api_key": "your-api-key"
}

Environment

  • Node.js 18.0.0 or higher

  • ZeroEntropy API access

License

MIT

Available Tools

1 tool
get_rerankingC

Get the reranked document listing

Inputs: query (required), documents (required), api_key (required).

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    C2.9/5.0

    Scored across 1 tool

    Disambiguation5/5

    With only one tool, there is no possibility of ambiguity or overlap between tools. The single tool 'get_reranking' has a clear and distinct purpose, so agents cannot misselect between non-existent alternatives.

    Naming Consistency5/5

    The naming pattern is trivially consistent as there is only one tool. It follows a verb_noun convention ('get_reranking'), which is clear and predictable, though no comparison with other tools is possible.

    Tool Count2/5

    A single tool is generally too few for most server purposes, as it limits functionality and suggests a thin or incomplete surface. For a 'Zerank' server, which might imply ranking or reranking operations, one tool feels insufficient to cover the domain adequately.

    Completeness2/5

    The tool set is severely incomplete for a reranking domain. With only a 'get_reranking' tool, there are obvious gaps such as creating, updating, or deleting rankings, or other related operations like listing or searching, which agents would need for full workflows.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      B
      quality
      C
      maintenance
      SEO intelligence tools for AI agents: SERPs, keyword research, backlinks, rank tracking, content planning, and site audits. Connect through hosted MCP or the open-source npm client. An AgentSEO account and API key are required; usage consumes credits.
      45
      90 npm
      MIT
    • A
      license
      Not graded
      quality
      D
      maintenance
      Enables AI agents to retrieve relevant sections from indexed files (documents, code, etc.) via hybrid semantic/BM25 search, with citations, through the Model Context Protocol.
      Apache 2.0