Prueba de protocolo Agent2Agent (A2A)

Las respuestas y explicaciones aparecen a medida que realiza el cuestionario.

1. ¿Cuál es el objetivo principal del protocolo A2A?
   • Training models on shared data
   • Comprimir pesos de modelos
   • Letting agents from different vendors discover each other and hand off tasks
   • Replacing HTTP
2. ¿Cómo se relaciona A2A con MCP?
   • MCP connects agents to tools and data, and A2A connects agents to other agents
   • A2A replaced MCP in 2025
   • They are the same protocol under two names
   • MCP is used only for image models
3. ¿Qué es una Tarjeta de Agente?
   • A payment method for agent usage
   • A hardware security key
   • A log of past conversations
   • A JSON document describing an agent's skills, endpoint and authentication
4. ¿En qué estándares se basa A2A?
   • FTP and XML-RPC
   • HTTP, JSON-RPC 2.0 and Server-Sent Events
   • Bluetooth and MQTT
   • SMTP and IMAP
5. ¿Qué significa el estado de tarea de "entrada requerida"?
   • The remote agent needs more information from the client before it can continue
   • The task finished successfully
   • The server is offline
   • The client must pay first
6. Why does A2A treat remote agents as opaque?
   • To make agents run faster
   • Because agents cannot use tools
   • So vendors can cooperate without exposing prompts, tools or models
   • Because JSON cannot describe internals
7. How can a client follow a task that takes hours?
   • It must keep polling every millisecond
   • A2A does not support long tasks
   • It has to restart the task repeatedly
   • By streaming progress or registering a webhook for push notifications
8. Why treat artifacts returned by another vendor's agent as untrusted?
   • They are always encrypted
   • They can contain prompt injection just as a web page can
   • A2A forbids reading them
   • They are always empty