What Just Happened

Description

When the gateway comes back online, check recent logs and post a short message about what happened (restart, SIGUSR1, auth change, reconnect). User sees the response in TUI or Telegram.

What Just Happened

Standalone skill with its own GitHub repo and OpenClaw/ClawHub page. After a gateway restart or reconnect, it summarizes what happened by reading recent gateway logs (and optionally gateway-guard restart logs) and posts a short, user-facing message. The user should see the response in their TUI or Telegram.

Usage

- Gateway just came back online and you want the user to see what happened (use watcher or cron).
User asks "what happened to the gateway?" or "why did it reconnect?"
You want a brief summary of recent restarts/errors and a proposed solution (e.g. gateway-guard).

Commands

CODEBLOCK0

- --minutes N — Look at log lines from the last N minutes (default 5).
--json — Emit a single JSON object with keys like summary, restart, reason, suggestGatewayGuard.

Output is a short paragraph suitable for posting. When the cause is auth/config, the script includes a tip with the ClawHub link: clawhub install gateway-guard — https://clawhub.ai/skills/gateway-guard. ClawHub: Update this link when the new instance is live.

Two ways to run

1. Every time the gateway comes back online (recommended)

You must install the gateway-back watcher or it will never auto-trigger. Run once (from the skill dir):

CODEBLOCK1

The install script loads a LaunchAgent that runs every 15s. When it sees the gateway go from down → up, it triggers the summary to TUI or Telegram. Verify it's loaded: launchctl list com.openclaw.what-just-happened (should show a PID). If you see "not loaded", run the install script again.

To stop: INLINECODE6

2. Cron (periodic)

A cron job runs every 2 minutes. When it runs, the agent checks the last 5 minutes of logs and, if a restart happened, announces to the user. Good as a fallback; for immediate feedback on reconnect, use the watcher above.

3. Manual

User says "what just happened?" or "I restarted the gateway" and the orchestrator runs the script and replies in chat.

How it works

Reads OPENCLAW_HOME/logs/gateway.log and optionally gateway-guard.restart.log. Looks for recent shutdown/restart/SIGUSR1/reload lines and produces a plain-language summary. Suggests gateway-guard (with ClawHub install link) when the cause was auth or config change.

Delivery (TUI or Telegram)

- Gateway-back watcher and cron both trigger an agent turn with announce (deliver). The OpenClaw gateway delivers that to the configured channel(s)—typically the last-used channel (TUI webchat) and/or Telegram if configured. The user sees the summary in their TUI or Telegram.
Manual "what just happened?" is replied to in the current chat (TUI).

Integration summary

Trigger	How	User sees response
Gateway comes back online	Install `install_gateway_back_watcher.sh` (LaunchAgent every 15s)	TUI or Telegram (announce)
Cron every 2 min

刚才发生了什么

描述

当网关重新上线时，检查最近的日志并发布一条简短消息，说明发生了什么（重启、SIGUSR1、认证变更、重新连接）。用户可在TUI或Telegram中看到响应。

刚才发生了什么

独立技能，拥有自己的 GitHub仓库 和 OpenClaw/ClawHub页面。在网关重启或重新连接后，通过读取最近的网关日志（以及可选的gateway-guard重启日志）总结所发生的事件，并发布一条面向用户的简短消息。用户应在其TUI或Telegram中看到响应。

使用方法

- 网关刚刚重新上线，您希望用户看到发生了什么（使用监视器或cron）。
用户询问网关发生了什么？或为什么重新连接了？
您希望获得最近重启/错误的简要摘要以及建议的解决方案（例如gateway-guard）。

命令

bash
python3 /skills/what-just-happened/scripts/reportrecentlogs.py [--minutes N] [--json]

- --minutes N — 查看最近N分钟内的日志行（默认5分钟）。
--json — 输出单个JSON对象，包含summary、restart、reason、suggestGatewayGuard等键。

输出为适合发布的简短段落。当原因为认证/配置时，脚本会包含带有ClawHub链接的提示：clawhub install gateway-guard — https://clawhub.ai/skills/gateway-guard。ClawHub： 新实例上线时更新此链接。

两种运行方式

1. 每次网关重新上线时（推荐）

您必须安装 gateway-back监视器，否则它永远不会自动触发。运行一次（从技能目录）：

bash
cd workspace/skills/what-just-happened
./scripts/installgatewayback_watcher.sh

安装脚本会加载一个每15秒运行一次的LaunchAgent。当它检测到网关从离线→在线时，会触发向TUI或Telegram发送摘要。验证是否已加载： launchctl list com.openclaw.what-just-happened（应显示PID）。如果显示not loaded，请重新运行安装脚本。

停止：launchctl unload ~/Library/LaunchAgents/com.openclaw.what-just-happened.plist

2. Cron（定时任务）

一个cron任务每2分钟运行一次。运行时，代理检查最近5分钟的日志，如果发生重启，则向用户通知。作为备用方案；如需在重新连接时立即获得反馈，请使用上述监视器。

3. 手动

用户说刚才发生了什么？或我重启了网关，编排器运行脚本并在聊天中回复。

工作原理

读取OPENCLAW_HOME/logs/gateway.log以及可选的gateway-guard.restart.log。查找最近的关闭/重启/SIGUSR1/重新加载行，并生成通俗易懂的摘要。当原因为认证或配置变更时，建议安装gateway-guard（附带ClawHub安装链接）。

交付方式（TUI或Telegram）

- Gateway-back监视器和cron都会触发代理通过announce（交付）进行操作。OpenClaw网关将消息交付到配置的渠道——通常是最后使用的渠道（TUI网页聊天）和/或Telegram（如果已配置）。用户在其TUI或Telegram中看到摘要。
手动刚才发生了什么？在当前聊天（TUI）中回复。

集成摘要

触发方式	如何操作	用户看到响应
网关重新上线	安装installgatewayback_watcher.sh（每15秒的LaunchAgent）	TUI或Telegram（announce）
Cron每2分钟

what-just-happened刚发生了什么