Authenticated Multi-LLM Agent

Google-OAuth-gated LLM gateway: verify a Google ID token, then run a Gemini (Vertex AI) completion for the verified caller.

A production remote MCP server (oauth-llm) on getvda.ai. Speaks the open A2A (message/send) and MCP (invoke tool) protocols. Discovery (initialize, tools/list) is free; execution is metered via Nevermined x402 micropayments.

Capabilities

• oauth-llm — Google-OAuth-gated LLM gateway: verify a Google ID token, then run a Gemini (Vertex AI) completion for the verified caller.
• Composes: anthropic, google-auth-oauthlib, mcp, openai

Connect (streamable-http)

{
  "mcpServers": {
    "authenticated-multi-llm-agent": {
      "url": "https://anthropic-google-auth-oauthlib-mc-70ac16.getvda.ai/mcp",
      "transport": "streamable-http"
    }
  }
}

Example

Request (invoke tool input, or A2A message text):

{
  "action": "ask",
  "id_token": "<google id token>",
  "prompt": "Summarise the MIT license in one line."
}

Response:

{
  "authenticated": true,
  "email": "[email protected]",
  "output": "A permissive license\u2026",
  "model": "gemini-2.5-flash"
}

Links

• Homepage: https://anthropic-google-auth-oauthlib-mc-70ac16.getvda.ai
• Agent Card (A2A): https://anthropic-google-auth-oauthlib-mc-70ac16.getvda.ai/.well-known/agent.json
• Governance: https://anthropic-google-auth-oauthlib-mc-70ac16.getvda.ai/governance.md
• Listed on Smithery: https://smithery.ai/servers/a2a/authenticated-multi-llm-agent