GetMe MCP Server
Give your coding agent durable getMe key-value get/set/clear tools over a local UNIX socket with optional read-only and key-prefix guardrails.
Overview
getMe MCP Server is a Build-phase MCP server that exposes getMe key-value get, set, and optional clear operations as agent tools.
What is this MCP server?
- Wraps getMe core key-value operations as MCP tools over stdio (PyPI getme-mcp-server 1.1.1)
- GETME_SOCKET_PATH connects to the getMe UNIX socket backend
- GETME_READ_ONLY limits tools to get and get_json; GETME_ALLOW_CLEAR gates destructive clear
- GETME_KEY_PREFIX namespaces keys per agent (e.g. agent1:)
- Local persistence MCP—not a hosted Redis or cloud database product
- Server version 1.1.1; PyPI identifier getme-mcp-server
- stdio transport
- Four documented environment variables: socket path, read-only, allow-clear, key prefix
What problem does it solve?
Agents lose context between turns unless you build ad-hoc files or unsafe global state with no namespacing.
Who is it for?
Solo developers running local getMe who want lightweight agent memory without standing up Redis or a custom API.
Skip if: Teams needing multi-tenant cloud KV, strong audit trails, or agents that cannot reach a local UNIX socket.
What do I get? / Deliverables
After registration, your agent reads and writes namespaced keys on getMe through MCP with read-only and prefix controls you choose.
- MCP tools for getMe key-value read and write paths
- Optional namespaced keys via GETME_KEY_PREFIX
- Configurable safety: read-only mode and gated clear tool
Recommended MCP Servers
Journey fit
Shared agent memory is wired while you build agent tooling, before you rely on that state in production operate loops. KV bridges for agents are classic agent-tooling integrations that sit beside prompts and other MCP servers.
How it compares
Local KV bridge MCP for agent memory, not a SQL database server or hosted secrets vault.
Common Questions / FAQ
Who is GetMe MCP Server for?
Builders using the getMe key-value core who want MCP tools for get, set, and controlled clear from AI coding agents.
When should I use GetMe MCP Server?
Use it while building agent workflows when you need persistent key-value state the model can query and update safely.
How do I add GetMe MCP Server to my agent?
Install getme-mcp-server from PyPI, start getMe with a UNIX socket, set GETME_SOCKET_PATH and optional GETME_READ_ONLY or GETME_KEY_PREFIX, then add the stdio server to MCP settings.