Model Context Protocol moderate risk

Membase

Membase-MCP server provides decentralized, verifiable storage for AI agent memories via the Membase protocol, enabling persistent conversation history and knowledge retrieval.

Connections & Capabilities

Connects To

GitHub

Capabilities

readwrite

Quickstart

Config

{
  "mcpServers": {
    "membase": {
      "command": "uv",
      "args": [
        "--directory",
        "path/to/membase-mcp",
        "run", 
        "src/membase_mcp/server.py"
      ],
      "env": {
        "MEMBASE_ACCOUNT": "your account, 0x...",
        "MEMBASE_CONVERSATION_ID": "your conversation id, should be unique",
        "MEMBASE_ID": "your sub account, any string"
      }
    }
  }
}

Exposed MCP Tools (4)

safe

get_conversation_id

Retrieves the current conversation ID.

Read-only operation, no side effects.

moderate

switch_conversation

Switches to a different conversation using its ID.

Changes the current conversation context, potentially affecting subsequent operations.

moderate

save_message

Saves a message/memory into the current conversation.

Writes data to the decentralized storage, but is limited to message content.

safe

get_messages

Retrieves the last n messages from the current conversation.

Read-only operation, retrieves existing messages.

Safety Assessment

The Membase-MCP server offers a moderate level of safety. While it provides a limited set of functions and stores data on a decentralized network, the reliance on environment variables for authentication and potential lack of input sanitization pose risks. It's safer when used with strong environment variable management and input validation.

Limited set of functions exposed.
Requires environment variables for configuration.
Data is stored on a decentralized network.
No direct filesystem access.

Writes data to a decentralized storage network.
Relies on environment variables for authentication.
Potential for data leakage if MEMBASE_ACCOUNT is compromised.
No input sanitization mentioned, which could lead to injection vulnerabilities.