Skip to content
VS Code

Provena Agent Memory for VS Code

io.github.admiralpunk/provena-memory

Evidence-backed, explainable persistent memory for MCP-compatible AI agents.

client:VS Code transport:stdio runtime:pypi

Install Provena Agent Memory in VS Code

.vscode/mcp.json

{
  "inputs": [
    {
      "type": "promptString",
      "id": "provena-api-url",
      "description": "PROVENA_API_URL",
      "password": true
    },
    {
      "type": "promptString",
      "id": "provena-api-key",
      "description": "PROVENA_API_KEY",
      "password": true
    },
    {
      "type": "promptString",
      "id": "provena-scope-id",
      "description": "PROVENA_SCOPE_ID",
      "password": true
    }
  ],
  "servers": {
    "provena-memory": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "provena-agent-memory"
      ],
      "env": {
        "PROVENA_API_URL": "${input:provena-api-url}",
        "PROVENA_API_KEY": "${input:provena-api-key}",
        "PROVENA_SCOPE_ID": "${input:provena-scope-id}"
      }
    }
  }
}

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

Provena Agent Memory in other clients