Synapse A2A is a communication framework that allows different AI agents (such as Claude, Gemini, and Codex) to exchange messages, share tasks, and coordinate workflows using the Google A2A protocol.

Can I interrupt an agent that is currently busy?

Yes, by sending a message with '--priority 5', you can issue an emergency interrupt to an agent that is in a 'PROCESSING' state.

How do I see all active agents in my project?

Run the 'synapse list' command to open a Rich TUI that displays the name, type, status, and port of every active agent in your current working directory.

How do I send a message to another agent using this skill?

You can use the 'synapse send "message"' command. You can also add the --response flag if you need the sender to wait for a reply from the target agent.

How does file safety work in a multi-agent environment?

The skill includes a 'file-safety' feature that allows agents to lock files they are currently working on, preventing other agents from making conflicting edits simultaneously.

Synapse A2A Agent Communication

Name: Synapse A2A Agent Communication
Author: s-hiraoku

bys-hiraoku

Collaboration Tools

Enables seamless inter-agent communication and coordination between Claude and other AI models using the A2A protocol.

About

Synapse A2A provides a comprehensive framework for multi-agent workflows, allowing Claude to communicate directly with other agents like Gemini, Codex, or Copilot. It manages complex message routing, automated response tracking, and priority levels while ensuring data integrity through built-in file-locking safety features. This skill is essential for developers building sophisticated multi-AI systems where agents must delegate tasks, share context, and synchronize operations within a unified environment.

Key Features

Real-time TUI for monitoring agent status (READY, WAITING, PROCESSING)
Searchable task history and persistent agent configuration management
Priority-based task execution with emergency interrupt support
File safety locking mechanism to prevent multi-agent write conflicts
0 GitHub stars
Bi-directional agent messaging with automated response tracking

Use Cases

Coordinating a multi-agent pipeline where one agent writes code and another performs automated reviews
Managing complex developer environments requiring simultaneous execution of multiple specialized AI agents
Synchronizing background tasks across different AI models while maintaining file system integrity

About

Key Features

Real-time TUI for monitoring agent status (READY, WAITING, PROCESSING)
Searchable task history and persistent agent configuration management
Priority-based task execution with emergency interrupt support
File safety locking mechanism to prevent multi-agent write conflicts
0 GitHub stars
Bi-directional agent messaging with automated response tracking

Use Cases

Coordinating a multi-agent pipeline where one agent writes code and another performs automated reviews
Managing complex developer environments requiring simultaneous execution of multiple specialized AI agents
Synchronizing background tasks across different AI models while maintaining file system integrity