360 Secure SkillHub
360 安全技能中心

OpenClaw Doctor

该 Skill 提供 OpenClaw 安装的全面健康诊断和自动修复，检查网关、系统资源、服务韧性、cron 自动化、频道连接、安全配置、内存与备份。支持 --fix 自动修复和 --quick 快速检查，主要面向 macOS 和 Linux，Windows 仅 WSL 基本支持。需要执行 bash/Node 脚本。

dagangtj
dagangtj
热度 38.60
最近更新 2026-05-11 17:27:20
分类：自动化依赖：代码执行 / 读写文件 / 第三方 API应用场景：工作流 / 自动化脚本 / DevOps / 报告撰写

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add openclaw-doctor
Metadata
name
openclaw-doctor
description
Comprehensive OpenClaw health diagnostics and auto-repair. Checks gateway, services, memory, cron, channels, disk, logs, and security. Auto-fixes common issues. Works on macOS and Linux. Use when: system feels slow, cron jobs fail, gateway won't start, channels disconnect, or for routine health checks.
version
1.0.0
author
管家 (AI Butler)

OpenClaw Doctor

A comprehensive health diagnostic and auto-repair skill for OpenClaw installations. Goes far beyond openclaw doctor with deep checks and automatic fixes.

When to Use

  • System feels slow or unresponsive
  • Cron jobs failing or not running
  • Gateway won't start or keeps crashing
  • Channels (Telegram/WhatsApp) disconnected
  • Routine health checks (add to heartbeat)
  • After system restart or power outage
  • Before and after upgrades

Quick Start

Ask your agent: "Run a full health check" or "Diagnose my OpenClaw setup"

The agent will execute scripts/diagnose.sh and interpret results.

What It Checks

1. Gateway Health

  • Process running and responsive
  • Port reachable and latency
  • LaunchAgent/systemd service status
  • KeepAlive and auto-restart configured
  • Auth token strength

2. System Resources

  • CPU load vs core count
  • Memory usage and swap pressure
  • Disk space (warn at 80%, critical at 90%)
  • Log file sizes (detect bloat)

3. Service Resilience

  • Auto-boot on power restore (macOS)
  • Auto-restart after crash
  • Sleep settings (servers should never sleep)
  • Node.js path validity (survives upgrades?)

4. Cron & Automation

  • All cron jobs status
  • Failed job detection
  • Stale jobs (not run in 2x their interval)
  • Session count and context usage

5. Channel Health

  • Each channel connectivity
  • Last message timestamps
  • Auth token validity

6. Security

  • Config file permissions (should be 600)
  • Gateway token strength
  • Exposed ports check
  • Workspace file permissions

7. Memory & Storage

  • Memory file count and total size
  • Stale daily logs (>30 days)
  • Backup status and recency
  • Vector index health

Auto-Fix Capabilities

The script can automatically fix:

  • Restart crashed gateway
  • Fix file permissions
  • Rotate oversized logs
  • Clear stale session data
  • Restart disconnected channels

Usage

# Full diagnostic (read-only)
bash scripts/diagnose.sh

# Diagnostic with auto-fix
bash scripts/diagnose.sh --fix

# Quick check (essential items only)
bash scripts/diagnose.sh --quick

Integration with Heartbeat

Add to your HEARTBEAT.md for periodic checks:

## System Health (every heartbeat)
Run: bash scripts/diagnose.sh --quick
Alert if any FAIL items found

Platform Support

  • macOS (primary, fully tested)
  • Linux (systemd-based, tested)
  • Windows WSL (basic support)