RE
OfficialDatabaseAI & ML

Redis Cloud API MCP Server for Natural Language

Manage Redis Cloud resources with natural language using the MCP server for Redis Cloud API to control and automate context between LLMs and your systems.

View on GitHub
Quick Install
npx -y @redis/mcp-redis-cloud

Overview

The Redis Cloud API MCP Server implements the Model Context Protocol (MCP) to let LLMs interact with Redis Cloud accounts through natural language. Run this server locally or in a container and connect it to any MCP-compatible client (for example, Claude Desktop or Cursor IDE) to perform account, subscription, and database operations using plain-language prompts.

This server maps high-level intents into calls against the Redis Cloud API and returns structured information and task IDs for long-running operations. It’s useful for automating routine administrative actions, exploring account state, and integrating Redis Cloud controls into conversational AI workflows while keeping the interaction model consistent with MCP clients.

Features

  • Account management: fetch current account metadata and payment methods
  • Subscription management:
    • Create, list, and delete Essential & Pro subscriptions
    • Configure multi-cloud, networking, persistence, modules, and Active-Active
  • Database capabilities: enumerate supported Redis modules and feature options
  • Cloud provider/region lookup: list available regions and zone options for AWS, GCP, Azure
  • Plans and pricing: query Essential and Flex plans (paginated)
  • Task tracking: inspect tasks and follow long-running operations by ID
  • Built for MCP: expose these operations as MCP tools consumable by LLM-based clients

Installation / Configuration

Prerequisites:

  • Node (tested with v22.14.0)
  • npm (tested with 10.x)
  • Optional: Docker

Clone the repo and install dependencies:

git clone https://github.com/redis/mcp-redis-cloud.git
cd mcp-redis-cloud
# Use nvm if needed:
nvm use v22.14.0
npm install
npm run build

Run locally:

# Provide credentials via env vars
API_KEY=<your_api_key> SECRET_KEY=<your_secret_key> node --experimental-fetch dist/index.js

Env vars

VariableDescription
API_KEYRedis Cloud API key
SECRET_KEYRedis Cloud API secret

Run with Docker:

# Build
docker build -t mcp/redis-cloud .

# Run
docker run --rm -e API_KEY=<your_api_key> -e SECRET_KEY=<your_secret_key> mcp/redis-cloud

Integrate with MCP clients (Claude Desktop example):

  1. Build the project (npm run build).
  2. Add an MCP server entry to your Claude Desktop config:
{
  "mcpServers": {
    "mcp-redis-cloud": {
      "command": "node",
      "args": ["--experimental-fetch", "/absolute/path/to/dist/index.js"],
      "env": {
        "API_KEY": "<redis_cloud_api_key>",
        "SECRET_KEY": "<redis_cloud_api_secret_key>"
      }
    }
  }
}

Restart the client to load the new server. Cursor IDE uses a similar mcp.json entry.

For inspecting tools during development:

npx @modelcontextprotocol/inspector node dist/index.js --api-key=<api_key> --secret-key=<secret_key>

Available Tools / Resources

The server exposes a set of MCP tools (actions) to the client. Representative tools include:

  • Accounts
    • get_current_account
    • get_current_payment_methods
  • Subscriptions (Pro & Essential)
    • get_pro_subscriptions
    • create_pro_subscription
    • get_essential_subscriptions
    • get_essential_subscription_by_id
    • create_essential_subscription
    • delete_essential_subscription
  • Database Capabilities
    • get_database_modules
  • Cloud Providers & Regions
    • get_pro_plans_regions
  • Plans & Pricing
    • get_essentials_plans
  • Task Management
    • get_tasks
    • get_task_by_id

Notes:

  • Many list endpoints are paginated — the client or user should iterate pages to retrieve full sets.
  • Long-running ops return task IDs which you can poll using task tools.

Use Cases

  • Create a new DB by natural language:

    • Prompt: “Create a 4GB Redis DB with persistence and RedisJSON in AWS us-east-1”
    • Server will map to create_pro_subscription/create_essential_subscription with appropriate options and return a task ID for tracking.

  • Inspect billing and payment methods:

    • Prompt: “Show my current Redis Cloud subscriptions and payment methods”
    • Server returns consolidated account and payment method data.

  • Choose the right plan for an application:

    • Prompt: “Recommend a Redis plan for an e-commerce product catalog (low latency, 50k ops/sec)”
    • Server can query available plans, regions, and features to inform the recommendation.

  • Monitor deployments:

    • After creating resources, poll task status:
    • Prompt: “What is the status of task <task_id>?” returns detailed progress and final state.

  • Multi-cloud deployments & migration planning:

    • Find supported modules and regions across AWS/GCP/Azure to plan Active-Active or cross-region topologies.

Development Tips

  • Rebuild after code changes:
npm run build
  • Test locally with the MCP Inspector (see earlier command).
  • Project layout highlights:
src/
├── index.ts          # server entry
├── clients/          # Redis Cloud API client code (generated)
└── tools/            # tool implementations (accounts, subscriptions, tasks)

This server is designed to be dropped into MCP-capable environments to extend LLM workflows with Redis Cloud operations in a secure, auditable way. For the full source and issues, see the repository: https://github.com/redis/mcp-redis-cloud/

Tags:ai-mldatabase