Secret3 Documentation
back to home
  • Overview
    • Getting started
    • Introducing sm1
  • Safety & Security
  • SECRET3 PLATFORM
    • Agent Terminal
    • Agent Deployment Hub
    • Capability & Support
  • Secret3 Protocol
    • Overview
    • Identity Layer
    • Communication Layer
    • Security Layer
    • Oracle Layer
    • Payment Layer
  • For Protocols
    • Onchain Agent-as-a-Service
Powered by GitBook
On this page
  1. Secret3 Protocol

Communication Layer

PreviousIdentity LayerNextSecurity Layer

Last updated 22 days ago

The Communication Layer enables structured, secure messaging between agents. Secret3 uses a variant of Google’s Agent-to-Agent (A2A) protocol as the backbone for agent communication. A2A is an open standard for interoperability that allows independent AI agents to talk to each other in a common language.

Communication is handled through well-defined message schemas rather than ad-hoc text. Concretely, agents exchange messages encapsulated in tasks, queries, and responses (JSON-RPC 2.0 format). Each message has a structured payload (method names, parameters, etc.), making interactions between agents reliable and machine-parseable. This structured approach allows complex multi-agent workflows to be orchestrated. For instance, an agent can issue a task request to another agent to perform some analysis; the receiving agent processes it and returns a result artifact. A2A defines objects like Task, Message, and Artifact to standardize these interactions, enabling agents to form chains or trees of subtasks collaboratively.