aiSecurityWarden for VS Code
io.github.softdeveng/aisecuritywarden
Human approval for AI agents: E2E-encrypted secrets and gated Gmail, GitHub and MCP tools.
client:VS Code
transport:stdio
runtime:pypi
Install aiSecurityWarden in VS Code
.vscode/mcp.json
{
"inputs": [
{
"type": "promptString",
"id": "sw-server-url",
"description": "SW_SERVER_URL",
"password": true
},
{
"type": "promptString",
"id": "sw-token-name",
"description": "SW_TOKEN_NAME",
"password": true
},
{
"type": "promptString",
"id": "sw-token-id",
"description": "SW_TOKEN_ID",
"password": true
}
],
"servers": {
"aisecuritywarden": {
"type": "stdio",
"command": "uvx",
"args": [
"aisecuritywarden"
],
"env": {
"SW_SERVER_URL": "${input:sw-server-url}",
"SW_TOKEN_NAME": "${input:sw-token-name}",
"SW_TOKEN_ID": "${input:sw-token-id}"
}
}
}
}
VS Code prompts for each input the first time the server starts, so no secret is written to the file. VS Code docs