mcp_server_memory

This MCP server implements a memory-focused MCP endpoint accessible over Server-Sent Events (SSE). The server is started by running the Python script with SSE transport on a chosen port. Before starting, you typically initialize the environment and activate the virtual environment, then launch the server. Once running, you can feed or build memory configuration data using the provided build workflow: memory_mcp.json is piped into build.py to create the in-memory MCP payload that the server will expose. The server exposes its MCP interface over SSE, allowing clients to subscribe to incremental updates and request or mutate memory-mapped configuration through the defined endpoints. Use the --port flag to choose the listening port and ensure it is accessible to clients that need to consume the MCP stream.

Prerequisites:

• Python 3.8+ (tested with a standard Python installation)
• Access to the repository containing mcp_server_memory

Installation steps:

1. Clone the repository (or download the package).
2. Create and activate a virtual environment:
   • On Unix/macOS: python3 -m venv .venv source .venv/bin/activate
   • On Windows (PowerShell): python -m venv .venv ..venv\Scripts\Activate.ps1
3. Install dependencies if provided (adjust as needed for your project):
   • If a requirements.txt exists: pip install -r requirements.txt
   • If dependencies are embedded, this step may be optional.
4. Initialize environment/state (if applicable): python init_env.py
5. Run the MCP server: python mcp_server.py --transport sse --port 8000
6. Build memory MCP payloads when needed: cat memory_mcp.json | python build.py (On Windows, use type memory_mcp.json | python build.py or Get-Content memory_mcp.json | python build.py)

Tips:

• The server uses Server-Sent Events (SSE); ensure clients support SSE subscriptions to receive real-time updates.
• Port 8000 is the default in the example; adjust if you need to avoid conflicts in your environment.
• memory_mcp.json is the source configuration for the in-memory MCP; use the provided build.py to transform it into the runtime payload.
• If you run into environment issues, ensure the virtual environment is activated and that Python can access the script paths.
• For Windows users, path separators and command syntax in the build step may differ (use PowerShell/cmd equivalents as shown).