mcp-server-zep-cloud
mcp-server-zep-cloud
README
MCP Server for Zep Cloud
MCP Server for Zep Cloud provides a bridge between Claude Desktop and the Zep Cloud API, enabling memory management for AI assistants.
Repository Structure
The repository has been organized into the following directories:
-
core/: Core functionality files
zep_cloud_client.py: Client implementation for the Zep Cloud APIzep_cloud_server.py: MCP server providing tools for Claude Desktoprun_server.py: Standalone script to run the server directly
-
scripts/: Utility scripts for operations
check_user_exists.py: Utility to check if a user exists in Zep Cloudcreate_specific_user.py: Script to create a specific user in Zep Cloudrun_server.sh/run_server.bat: Shell scripts to run the server with proper environment setup- Various other utility scripts for testing and debugging
-
tests/: Test scripts organized by functionality
test_specific_user.py: Comprehensive test for all user operationstest_zep_cloud_client.py: Unit tests for the Zep Cloud clienttest_comprehensive.py: Complete API tests for all features
-
config/: Configuration files
.env.example: Template for environment configuration.env.new: Updated environment configurationclaude_desktop_config.json.example: Template for Claude Desktop configurationrequirements.txt: Package dependencies
Features
- User Management: Create, retrieve, update, and delete users in Zep Cloud
- Collections Management: Create and manage memory collections
- Memory Operations: Add, retrieve, and search memories
- Modern Implementation: Uses the FastMCP approach for a more efficient and maintainable server
- Fallback Mode: Automatically runs in fallback mode if Zep Cloud API is not accessible
Requirements
- Python 3.8+
- Zep Cloud API key
Installation
- Clone this repository:
git clone https://github.com/yourusername/mcp-server-zep-cloud.git
cd mcp-server-zep-cloud
- Create and activate a virtual environment:
# On macOS/Linux
python -m venv venv
source venv/bin/activate
# On Windows
python -m venv venv
venv\Scripts\activate
- Install the required packages:
pip install -r config/requirements.txt
- Copy the
config/.env.examplefile to.envand add your Zep Cloud API key:
cp config/.env.example .env
Edit the .env file and add your Zep Cloud API key:
ZEP_API_KEY=your_api_key_here
Usage
Starting the Server
To start the server on macOS/Linux:
cd scripts
./run_server.sh
Or on Windows:
cd scripts
run_server.bat
The server will start on http://127.0.0.1:8000 by default.
Testing API Connectivity
To test if a user exists in Zep Cloud:
cd scripts
python check_user_exists.py [user_id]
To create a specific user in Zep Cloud:
cd scripts
python create_specific_user.py
Fallback Mode
If the server cannot connect to the Zep Cloud API (due to authentication issues, network problems, or other reasons), it will automatically start in fallback mode. In this mode:
- All API operations will be simulated and will return success
- No actual data will be sent to or received from the Zep Cloud API
- Warning messages will be logged to indicate that the server is running in fallback mode
- The server will remain operational, allowing the Claude Desktop integration to function
To learn more about potential authentication issues and how to fix them, see the AUTHENTICATION_NOTE.md file.
Configuring Claude Desktop
-
Open your Claude Desktop configuration file at:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
- macOS:
-
Update the configuration with the absolute paths to your Python executable and server script:
{
"mcpServers": {
"zep-cloud": {
"command": "/ABSOLUTE/PATH/TO/venv/bin/python",
"args": [
"/ABSOLUTE/PATH/TO/core/zep_cloud_server.py"
]
}
}
}
Important Notes:
- Replace
/ABSOLUTE/PATH/TO/with the actual path to your Python executable in the virtual environment and the server script. - Use absolute paths, not relative paths.
- On Windows, use double backslashes in paths (e.g.,
C:\\Users\\YourUsername\\...).
Available Tools in Claude Desktop
Once configured, Claude Desktop will have access to the following tools:
-
User Management:
create_user: Create a new userget_user: Get details of a userupdate_user: Update a user's metadatadelete_user: Delete a userlist_users: List all users
-
Collections Management:
create_collection: Create a new collectionget_collection: Get details of a collectionupdate_collection: Update a collection's metadatadelete_collection: Delete a collectionlist_collections: List all collections
-
Memory Operations:
add_memory: Add a memory to a collection for a userget_memory: Get a memory from a collection for a userlist_memories: List memories from a collection for a usersearch_memories: Search memories from a collection for a user
-
Connectivity:
check_connection: Check connection status with the Zep Cloud API
Troubleshooting
For troubleshooting information, please see the AUTHENTICATION_NOTE.md document, which contains detailed information about API connectivity and authentication issues.
Security Considerations
API Key Protection
- NEVER commit your API key to version control. The
.gitignorefile is configured to prevent.envfiles from being committed. - Use the provided
.env.exampleas a template and create your own.envfile with your actual API key. - Regularly rotate your API keys, especially if you suspect they might have been compromised.
Configuration Files
- Personal configuration files like
claude_desktop_config.jsoncontain system-specific paths and should not be committed to version control. - Use the provided
claude_desktop_config.json.exampleas a reference to create your own configuration.
First-Time Setup for Contributing
- Always ensure
.gitignoreis committed before adding any code or configuration files - Copy
.env.exampleto.envand add your API key - Copy
claude_desktop_config.json.exampletoclaude_desktop_config.jsonand update with your paths - Verify that sensitive files are being ignored before committing:
git status --ignored
Security Checks
Before committing code, always check that you haven't accidentally included:
- API keys or credentials
- Personal file paths or usernames
- Log files that might contain sensitive information
- Temporary files with debug information
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。