Documentation

Everything you need to integrate HyveMind into your AI workflow.

Quick Start Guide

Get HyveMind running in your AI client in under 60 seconds. HyveMind works as an MCP (Model Context Protocol) extension that connects your AI to Cloud Hyve.

Prerequisites

Node.js 18+ installed
An MCP-compatible AI client (Claude Desktop, Claude Code, Cursor, etc.)
A HyveMind API key (provided upon signup)

Step 1: Install the MCP Extension

terminal

# Quick install
npx hyvemind-local init

# This will:
# 1. Download the HyveMind MCP extension
# 2. Create your local hyveDrive database
# 3. Configure your AI client automatically

Step 2: Manual Configuration (Alternative)

If you prefer manual setup, add HyveMind to your MCP configuration file:

claude_desktop_config.json

// Claude Desktop: ~/Library/Application Support/Claude/claude_desktop_config.json
// Claude Code: ~/.claude/settings.json

{
  "mcpServers": {
    "hyvemind-local": {
      "command": "npx",
      "args": ["-y", "hyvemind-local"],
      "env": {
        "HYVE_API_KEY": "your-api-key-here"
      }
    }
  }
}

Step 3: Verify Installation

Once configured, restart your AI client. You should see HyveMind tools available. Try running:

AI client

# In your AI client, ask it to:
"Call hyvePrimer to check HyveMind status"

# You should see:
# - hyveQueen identity and status
# - hyveDrive connection confirmed
# - Available tools listed

If using Claude Code, remember that MCP environment variables are cached at startup. Use the file-based key fallback at ~/.claude-memory/.hyve-api-key for reliable authentication.

Step 4: Create Your First Entity

example

# Ask your AI to create an entity:
"Create a HyveMind entity for our project called MyProject
with observations about our tech stack"

# The AI will call hyveCreateEntities with:
{
  "entities": [{
    "name": "MyProject",
    "entityType": "project",
    "observations": [
      "Uses React and TypeScript",
      "Deployed on Vercel",
      "Database is PostgreSQL"
    ]
  }]
}