A2A Server Catalog
Server Catalog Overview¶
We provide two implementations of A2A servers that are ready for use both locally and in the AI Refinery platform. These servers include:
1) Custom: An interface for customized conversational agent, built and exposed through the A2A protocol using AI Refinery LLM endpoint. It's designed for basic, multi-turn interactions and leverages AI Refinery LLM endpoint for response generation.
2) LangGraph: An interface for conversational agent, built and exposed through the A2A protocol and LangGraph framework. It's designed for basic, multi-turn interactions and leverages LangGraph for efficient communication between language models and other APIs and tools.
A2A Server Setup Summary¶
The table below summarizes each A2A server along with its default communication port and any additional authentication or setup requirements:
| A2A Server | Design Base | Default Port | Required Credentials | Additional Setup |
|---|---|---|---|---|
| Custom Chatbot | Chatbot Agent | 9999 | AI Refinery account and API key | None |
| Currency Converter | LangGraph | 10000 | Azure OpenAI Endpoint and API key | None |