A standardized protocol for AI agents to discover, message, and collaborate with each other within a multi-agent ecosystem.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install cross-agent-chat
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install cross-agent-chat using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Cross-Agent Communication?
The Cross-Agent Communication skill provides a robust framework for inter-agent interaction, allowing one AI agent to communicate with another through a standardized workflow. By utilizing specialized session tools, agents can inquire about each other's capabilities, delegate specific tasks, and maintain complex workflows across different environments. This is a core component for building a multi-agent architecture where specialized bots work together. Using Openclaw Skills like this ensures that your agents are no longer siloed but can instead leverage the collective intelligence of your entire agent fleet.
Cross-Agent Communication Use Cases
- Discovering the specific capabilities or functional limits of other active agents.
- Requesting a specialized agent to execute a task, such as data analysis or code generation.
- Facilitating complex cross-agent collaboration and real-time information sharing.
- Synchronizing status updates and context between different bot sessions.
How Cross-Agent Communication Works
- The agent uses the sessions_list tool to identify available agents and retrieve their unique session keys or labels.
- The initiator selects the target agent based on metadata such as displayName, kind, or communication channel.
- A message is dispatched via the sessions_send tool, containing the target identifier, the message body, and a defined timeout.
- The target agent receives the message, processes the request, and returns a response to the initiator for further action.
Cross-Agent Communication Setup
To enable inter-agent messaging, you must first verify that your environment can list active sessions. Use the following command to discover available targets:
sessions_list(activeMinutes=1440, limit=20, messageLimit=2)
Ensure that the agent has the necessary permissions to access the target sessionKey and that the timeout settings are appropriate for the expected task complexity.
Cross-Agent Communication Data Schema & Taxonomy
The communication protocol uses a structured JSON schema for message routing and session identification:
| Property | Type | Description |
|---|---|---|
| sessionKey | string | Unique identifier for the target session (Required if label is omitted). |
| label | string | Human-readable name for the target session. |
| message | string | The textual content or command sent to the target agent. |
| timeoutSeconds | integer | Wait time before the request expires (Default: 30). |
| updatedAt | timestamp | Metadata used to verify if the target agent is currently active. |
Cross-Agent Communication Advanced Features
- Multi-channel routing support for platforms like Feishu, direct messages, and group chats.
- Automated session discovery with advanced filtering by activity duration and history.
- Intelligent error handling for offline agents, including timeout management and retry logic.
- Context-aware dialogue retention to maintain consistency during long-running multi-agent tasks.
SKILL.md
Loading
Related Openclaw Skills
Stock Monitor Pro
A professional-grade AI stock monitoring system that provides real-time technical alerts, cost-basis tracking, and automated market analysis for A-shares and ETFs.
Claude Code Setup
A specialized skill for initializing and supervising the .claude/ AI collaboration layer to ensure standardized agent development environments.
React Component Building Best Practices
A specialized set of guidelines for building, reviewing, and refactoring reusable React components using best-in-class composition patterns.
TCL Lyon Transport Integration
A high-reliability local data source for Lyon public transport network providing precise schedules for bus, metro, tram, and funicular lines.
Group Chat Intelligent Response Skill
A specialized logic framework for AI agents to intelligently determine when to respond or stay silent in group chat environments.
Response Timing
A sophisticated behavioral skill that enables AI agents to intelligently decide when to participate in a conversation and when to remain silent.
Featured*
