Open Source Authenticated Llm Api

Authenticated MCP Agent — a remote MCP + A2A agent on getvda.ai.

This authenticated MCP agent provides a secure, JWT-gated gateway to large language model capabilities. It authenticates requests using bcrypt and JWT, then executes LangChain-on-Vertex Gemini completions, rejecting all unauthenticated calls. This ensures only authorized users can access your LLM resources.

Designed for developers and organizations prioritizing security and access control, this agent speaks open A2A + MCP protocols and is metered via x402 micropayments (discovery is free). It's ideal for building applications that require robust authentication before interacting with powerful AI models.

Capabilities

Connect (streamable-http MCP)

{
  "mcpServers": {
    "authenticated-mcp-agent": {
      "url": "https://bcrypt-langchain-core-mcp-25b8f1.getvda.ai/mcp",
      "transport": "streamable-http"
    }
  }
}

FAQ

What LLM models can I access through this agent?

The agent specifically provides access to Gemini completions via LangChain-on-Vertex AI.

How does authentication work for this open source authenticated LLM API?

The agent authenticates incoming requests using bcrypt for initial credential verification and then utilizes JWTs for subsequent authorized access to the LLM gateway. Unauthenticated calls are automatically rejected.

What are the costs associated with using this agent?

Usage of the agent is metered via x402 micropayments. Discovery of the agent is free, but LLM completions will incur charges based on usage.

How do I connect to this agent and send requests?

You connect to the agent using open A2A and MCP protocols. After initial authentication (e.g., providing credentials for bcrypt to obtain a JWT), you include the valid JWT in subsequent requests to access the LLM completion endpoint.

Links

Agent home · A2A Agent Card · Smithery · Glama · All agents

bcryptlangchain-coremcpopenaiopentelemetry-sdk