Back to MCP directory
publicPublicdnsLocal runtime

memory-bank-mcp-server

通过全局和分支特定的记忆库,管理跨 Claude AI 会话的项目文档和上下文的服务器,通过结构化的 JSON 文档存储实现一致的知识管理。

article

README

Memory Bank MCP Server

A Memory Bank implementation for managing project documentation and context across sessions using the Model Context Protocol (MCP). This server helps AI agents like Claude maintain consistent project knowledge through global and branch-specific memory banks stored in a structured JSON format.

This project is inspired by Cline Memory Bank from the nickbaumann98/cline_docs repository.

Packages

This repository is a monorepo managed with Yarn Workspaces. It contains the following packages:

  • packages/mcp: The core MCP server implementation. Contains the main logic for handling memory bank operations, MCP tool execution, and server startup.
  • packages/schemas: Defines the JSON schemas used for memory bank documents (e.g., memory_document_v2).
  • packages/vscode-extension: A VSCode extension providing integration with the Memory Bank MCP server (details TBD).

Getting Started

Prerequisites

  • Node.js (see .tool-versions for recommended version)
  • Yarn (v1.x)

Installation

Clone the repository and install dependencies from the root directory:

git clone https://github.com/t3ta/memory-bank-mcp-server.git
cd memory-bank-mcp-server
yarn install

Running the MCP Server

You can run the MCP server directly from the monorepo:

# From the monorepo root directory
yarn workspace @memory-bank/mcp start --docs /path/to/your/docs

Replace /path/to/your/docs with the actual path to your project's documentation directory (where global-memory-bank and branch-memory-bank will reside or be created).

See the packages/mcp/README.md for more details on running the server and its options.

Development

  • Build all packages: yarn build
  • Run tests for all packages: yarn test
  • Lint code: yarn lint

Refer to the README file within each package directory for package-specific development instructions.

License

This project is licensed under the MIT License. See the LICENSE file for details.

help

Runtime guide

cloud

Hosted runtime

Hosted servers run from a provider-managed environment. You usually connect the MCP client to the hosted endpoint or follow the provider's authorization flow, without keeping a local process alive

  1. Open provider connection page
  2. Authorize or copy endpoint
  3. Connect from your MCP client
terminal

Local runtime / other methods

Local servers run on your own machine or infrastructure. You normally copy the server_config into your MCP client, install the required package, and provide env variables from env_schema when needed

  1. Copy server_config
  2. Install required package
  3. Fill env variables and restart client