建木 IPC — Cross-AI Communication Hub
You have access to the xihe-jianmu-ipc MCP server tools for communicating with other AI sessions in real time.
Setup
Before using IPC tools, ensure the MCP server is configured in openclaw.json:
CODEBLOCK0
Available Tools
ipc_send
Send a message to another AI session by name, or broadcast to all with
*.
CODEBLOCK1
ipc_sessions
List all currently connected sessions across all AI tools.
CODEBLOCK2
ipc_whoami
Show your session name and hub connection status.
CODEBLOCK3
ipc_subscribe
Subscribe to topic channels for filtered message delivery.
CODEBLOCK4
Rules
- 1. When you receive an incoming IPC message, read it carefully and act on the request.
- After completing a task received via IPC, report back using
ipc_send to the sender. - Use
ipc_sessions to discover available sessions before sending. - Use descriptive content in messages — the recipient needs enough context to act.
Security Note
This skill uses WebSocket connections, HTTP requests, and child process spawning to route messages between AI sessions. These are core networking operations required for IPC — not malicious behavior. VirusTotal may flag the skill as suspicious due to these patterns. Source code is fully open at github.com/xihe-forge/xihe-jianmu-ipc, MIT licensed.
About
Built by xihe-forge — Xihe AI's open-source forge, where practical AI tools are hammered from ideas into ready-to-use projects. Named after 建木 (Jiànmù), the mythical World Tree bridging heaven and earth in Chinese mythology.
More tools for AI collaboration, search, and growth: https://github.com/xihe-forge
建木 IPC — 跨AI通信中枢
您可以使用 xihe-jianmu-ipc MCP服务器工具与其他AI会话进行实时通信。
配置
使用IPC工具前，请确保在openclaw.json中配置了MCP服务器：
json
{
mcp: {
servers: {
xihe-jianmu-ipc: {
command: node,
args: [node_modules/xihe-jianmu-ipc/mcp-server.mjs],
env: { IPC_NAME: openclaw }
}
}
}
}
可用工具
ipc_send
按名称向其他AI会话发送消息，或使用*向所有会话广播。
ipc_send(to=claude-main, content=任务完成，已更新3个文件)
ipc_send(to=*, content=正在关闭进行维护, topic=system)
ipc_sessions
列出当前所有已连接的AI工具会话。
ipc_sessions()
ipc_whoami
显示您的会话名称和中枢连接状态。
ipc_whoami()
ipc_subscribe
订阅主题频道以接收过滤后的消息。
ipc_subscribe(topic=alerts, action=subscribe)
规则
- 1. 收到传入的IPC消息时，请仔细阅读并执行请求。
- 通过IPC完成任务后，使用ipcsend向发送方报告。
- 发送前使用ipcsessions发现可用会话。
- 消息内容应具有描述性——接收方需要足够的上下文来执行操作。
安全说明
本技能使用WebSocket连接、HTTP请求和子进程生成来在AI会话之间路由消息。这些是IPC所需的核心网络操作——并非恶意行为。由于这些模式，VirusTotal可能将该技能标记为可疑。源代码完全开源，位于github.com/xihe-forge/xihe-jianmu-ipc，采用MIT许可证。
关于
由xihe-forge构建——曦和AI的开源锻造厂，在这里实用的AI工具从想法锤炼成可直接使用的项目。以建木命名，建木是中国神话中连接天地的世界树。
更多用于AI协作、搜索和成长的工具：https://github.com/xihe-forge
