Graph-Mem for VS Code

io.github.sathvik-1007/graphmem-mcp

Persistent knowledge-graph memory for AI agents, with local semantic search. No API keys.

client:VS Code transport:stdio runtime:pypi

Install Graph-Mem in VS Code

.vscode/mcp.json

{
  "inputs": [
    {
      "type": "promptString",
      "id": "graphmem-db-path",
      "description": "GRAPHMEM_DB_PATH",
      "password": true
    },
    {
      "type": "promptString",
      "id": "graphmem-embedding-model",
      "description": "GRAPHMEM_EMBEDDING_MODEL",
      "password": true
    },
    {
      "type": "promptString",
      "id": "graphmem-traversal-node-budget",
      "description": "GRAPHMEM_TRAVERSAL_NODE_BUDGET",
      "password": true
    },
    {
      "type": "promptString",
      "id": "graphmem-log-level",
      "description": "GRAPHMEM_LOG_LEVEL",
      "password": true
    }
  ],
  "servers": {
    "graphmem-mcp": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "graphmem-mcp"
      ],
      "env": {
        "GRAPHMEM_DB_PATH": "${input:graphmem-db-path}",
        "GRAPHMEM_EMBEDDING_MODEL": "${input:graphmem-embedding-model}",
        "GRAPHMEM_TRAVERSAL_NODE_BUDGET": "${input:graphmem-traversal-node-budget}",
        "GRAPHMEM_LOG_LEVEL": "${input:graphmem-log-level}"
      }
    }
  }
}

VS Code prompts for each input the first time the server starts, so no secret is written to the file. VS Code docs

Graph-Mem in other clients

Graph-Mem MCP server