AWS Labs Memcached MCP Server

raw JSON →
1.0.18 verified Sat May 09 auth: no python

An AWS Labs Model Context Protocol (MCP) server for Amazon ElastiCache Memcached. Current version 1.0.18, requires Python >=3.10. Provides tools to interact with Memcached via MCP. Active development with frequent releases.

pip install awslabs-memcached-mcp-server
error ModuleNotFoundError: No module named 'awslabs_memcached_mcp_server'
cause Trying to import using the PyPI name with dashes.
fix
Use import memcached_mcp_server (underscores, no 'awslabs-' prefix).
error ImportError: cannot import name 'main' from 'memcached_mcp_server'
cause Outdated version or wrong import statement.
fix
Ensure version >=1.0.18 and use: from memcached_mcp_server import main.
error pkg_resources.VersionConflict: mcp 0.x.x does not satisfy 1.x.x
cause Conflicting MCP package version in environment.
fix
Create a fresh virtual environment and reinstall: pip install awslabs-memcached-mcp-server.
breaking The package is named `awslabs-memcached-mcp-server` on PyPI but the import path is `memcached_mcp_server`. Using dashes in imports will fail.
fix Use `import memcached_mcp_server` or `from memcached_mcp_server import main`.
gotcha Requires Python >=3.10. Running on older Python will cause installation failure.
fix Upgrade Python to 3.10 or later.
gotcha Default endpoint is localhost:11211; does not automatically discover AWS ElastiCache clusters unless configured via environment variables.
fix Set MEMCACHED_ENDPOINT env var to your ElastiCache endpoint, or pass it as a parameter to main().

Run the MCP server with Memcached endpoint from environment variable MEMCACHED_ENDPOINT (default localhost:11211).

from memcached_mcp_server import main
from dotenv import load_dotenv
import os

load_dotenv()
# Set endpoint via env var or parameter
os.environ.get('MEMCACHED_ENDPOINT', 'localhost:11211')
main()