SmartMemory

Provides memory-enabled AI capabilities by converting conversations into a verified knowledge graph with rule learning.

python

3

GitHub Stars

python

Language

2 months ago

First Indexed

3 weeks ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "mauriceisrael-smartmemory": {
      "command": "/absolute/path/to/SmartMemory/venv/bin/python",
      "args": [
        "-m",
        "smart_memory.server"
      ]
    }
  }
}

SmartMemory is an MCP server that enables your AI assistants to remember facts, learn business rules, and derive new insights by interacting with a structured knowledge graph. It supports conversational memory and rule extraction workflows, making it easier to transform dialogue into verifiable knowledge for your AI workflows.

How to use

You use SmartMemory by connecting an MCP client to either the conversational memory mode or the supervision workflow. In conversational mode, your assistant gains long-term memory and logical deduction from chat sessions. In supervision mode, you deploy a dashboard-backed server to visualize the knowledge graph and extract rules from documents.

To start in conversational mode, configure the MCP client to connect to the SmartMemory MCP server using the provided command settings. Once connected, your assistant can remember facts, learn rules from natural dialogue, and apply those rules in subsequent conversations.

To start in supervision mode, deploy the full dashboard via Docker or run a local server. This mode lets you upload documents, configure the LLM, and review proposed rules in a visual interface.

How to install

Prerequisites you should have before installation: Python 3.11 or newer, Docker if you want to run the dashboard, and a compatible MCP client for your AI assistant.

Option 1 — Docker (Recommended for most users) to run the conversational memory server directly inside a container.

Option 2 — Local server (Private) for developers and privacy-conscious users who want to run from source.

Proceed with the following steps depending on your chosen path.

Additional sections

Configuration notes: SmartMemory exposes two primary modes. The conversational mode connects your MCP client to a memory-enabled brain, while the supervision mode runs a web dashboard and API server to manage the knowledge graph and extraction workflows.

Security considerations: When running in private or shared environments, restrict access to the dashboard and use secure keys for any external LLM providers. Prefer ephemeral containers for single-use environments and rotate credentials regularly.

Examples of typical usage: start a conversation with the assistant, say a fact like “Bob goes to work by car.” then ask questions such as “Can Bob vote?” The system can propose rules like “Drivers are adults,” and you can approve them to expand your knowledge graph.

Troubleshooting tips: If the MCP client cannot connect, verify the server is reachable on the expected port, check that the LLM provider configuration is valid, and ensure the necessary language model or local LLM binaries are accessible.

Available tools

interactive_demo

Demonstrates how the conversational memory flow captures facts, proposes rules, and confirms rule creation in an interactive session.

dashboard

Web dashboard to visualize the knowledge graph, review extracted rules, and manage documents for rule extraction.

extract_rules

Processes documents (e.g., PDFs) to propose business rules from their content, which you review and approve in bulk or individually.