Open Source Authenticated Llm Api

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

The Authenticated MCP Traced Agent provides a robust and secure open source authenticated LLM API. It acts as a JWT-gated gateway, ensuring all Large Language Model (LLM) completion requests are authenticated via bcrypt and JWT before execution on a LangChain-on-Vertex Gemini backend. Unauthenticated calls are automatically rejected.

Designed for developers requiring secure, auditable, and interoperable LLM access, this agent speaks open A2A and MCP protocols. It's metered via x402 micropayments, offering discovery free access and transparent usage billing for your AI interactions.

Capabilities

Connect (streamable-http MCP)

{
  "mcpServers": {
    "authenticated-mcp-traced-agent": {
      "url": "https://bcrypt-cryptography-langchain-cor-027cf2.getvda.ai/mcp",
      "transport": "streamable-http"
    }
  }
}

FAQ

How does this agent handle authentication?

The agent uses bcrypt for password hashing and JSON Web Tokens (JWT) for session management. All incoming LLM completion requests must present a valid JWT; otherwise, they are rejected.

What LLM models can I access through this gateway?

This agent is specifically configured to run LangChain-on-Vertex Gemini completions, providing access to Google's Gemini models via your Vertex AI setup.

How is usage priced or metered?

The agent is metered via x402 micropayments, meaning you pay for actual usage. Discovery of the service is free, and costs are incurred per LLM completion request.

How do I connect to this authenticated LLM API?

You connect to the agent's API endpoint by including a valid JWT in your request headers. The agent will then authenticate your request and forward it to the LangChain-on-Vertex Gemini backend for completion.

Links

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

bcryptcryptographylangchain-coremcpopentelemetry-sdkpython-jose