Your own AI agent, live in a Google Meet or Zoom call: reads the captions, answers in a side panel, runs entirely on your machine over MCP.
Setup
Choose your client, copy the config, then fill in any required environment variables.
claude mcp add meet-live-assist --env MLA_TOKEN='<MLA_TOKEN>' -- npx -y meet-live-assist-server~/.codex/config.toml
[mcp_servers.meet-live-assist]
command = "npx"
args = ["-y", "meet-live-assist-server"]
env = { MLA_TOKEN = "<MLA_TOKEN>" }
# codex mcp add meet-live-assist --env MLA_TOKEN='<MLA_TOKEN>' -- npx -y meet-live-assist-server~/.cursor/mcp.json
{
"mcpServers": {
"meet-live-assist": {
"command": "npx",
"args": [
"-y",
"meet-live-assist-server"
],
"env": {
"MLA_TOKEN": "<MLA_TOKEN>"
}
}
}
}.vscode/mcp.json
{
"servers": {
"meet-live-assist": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"meet-live-assist-server"
],
"env": {
"MLA_TOKEN": "<MLA_TOKEN>"
}
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"meet-live-assist": {
"command": "npx",
"args": [
"-y",
"meet-live-assist-server"
],
"env": {
"MLA_TOKEN": "<MLA_TOKEN>"
}
}
}
}claude_desktop_config.json
{
"mcpServers": {
"meet-live-assist": {
"command": "npx",
"args": [
"-y",
"meet-live-assist-server"
],
"env": {
"MLA_TOKEN": "<MLA_TOKEN>"
}
}
}
}.trae/mcp.json
{
"mcpServers": {
"meet-live-assist": {
"command": "npx",
"args": [
"-y",
"meet-live-assist-server"
],
"env": {
"MLA_TOKEN": "<MLA_TOKEN>"
}
}
}
}Environment variables
MLA_TOKEN
Replace the placeholders with your own credentials and keep them out of version control.