← Back to Skills Marketplace
doradx

OpenClaw Watchdog Pro

by Dorad · GitHub ↗ · v2.2.0 · MIT-0
cross-platform ⚠ suspicious
173
Downloads
0
Stars
0
Active Installs
3
Versions
Install in OpenClaw
/install openclaw-watchdog-pro
Description
OpenClaw 跨平台配置备份与网关监控。自动备份 openclaw.json,每分钟检查 gateway 状态,宕机时自动恢复。支持 Linux/macOS/Windows。触发:配置备份、gateway 监控、自动恢复、看门狗、watchdog。
Usage Guidance
This skill appears to implement the advertised backup and gateway-monitoring features, but it is intrusive and requires elevated installation steps. Before installing: (1) Review the bundled scripts yourself — they will create systemd/launchd/crontab/schtasks entries and add a shell alias. (2) Note the shell script uses /root/.openclaw (inconsistent with the Node scripts) — this could cause files to be written under root unexpectedly. (3) Installing the systemd/Windows task will run the watchdog as root/SYSTEM (high privilege); only proceed if you trust the source and the openclaw commands invoked (start/stop/doctor). (4) If uncertain, run the scripts in a disposable VM or container first, or manually inspect/copy only the parts you need (for example, avoid enabling the system service and run the monitor in a user session). (5) Correct the minor inconsistencies (install.js vs install.cjs) or request clarifications from the publisher. If you rely on sensitive host integrity, treat this as high-risk until vetted.
Capability Analysis
Type: OpenClaw Skill Name: openclaw-watchdog-pro Version: 2.2.0 The bundle implements a cross-platform system watchdog that establishes high-privilege persistence via systemd (running as root), Windows Task Scheduler (running as SYSTEM), and macOS launchd. It uses `execSync` in `watchdog.cjs` and `watchdog.sh` to wrap the main `openclaw` CLI, which presents a risk of command injection if inputs are not properly handled. While these behaviors are high-risk and provide persistent system-level execution, they are functionally aligned with the stated purpose of monitoring and auto-recovering a gateway service. No evidence of data exfiltration, obfuscation, or intentional backdoors was found.
Capability Assessment
Purpose & Capability
Name/description (backup + gateway monitoring) align with the scripts: they back up openclaw.json, probe the gateway via WebSocket/TCP, attempt recoveries, and call openclaw doctor. However there are minor mismatches (SKILL.md references install.js but file is install.cjs) and the shell script hardcodes /root/.openclaw while the Node scripts use the user's home or OPENCLAW_CONFIG_DIR — this inconsistency is unexpected for a cross-platform user-space tool.
Instruction Scope
Runtime instructions and code modify system state: install script writes systemd unit files, edits user shell rc (~/.bashrc, ~/.profile), adds cron tasks or Windows scheduled tasks, and instructs aliasing oc to wrap openclaw. The watchdog will execute system commands (openclaw start/stop/doctor) and probe localhost ports. SKILL.md does not explicitly warn about needing root for systemd/schtasks changes. The shell script hardcodes /root/.openclaw which could cause unexpected behavior or elevated-file use when run as non-root or root.
Install Mechanism
No remote downloads are performed (no install spec); code is bundled with the skill. That lowers supply-chain risk, but the included install script writes to privileged locations (/etc/systemd/system) and calls systemctl/schtasks/launchctl. The install behavior requires running the bundled script with sufficient privileges to create services/tasks; this is expected but intrusive.
Credentials
The skill does not request environment variables or external credentials. It uses OPENCLAW_CONFIG_DIR (optional) and otherwise operates on local files and runs local openclaw commands — consistent with its purpose.
Persistence & Privilege
Installer configures persistent, privileged runners: systemd service created at /etc/systemd/system (service runs as User=root), launchd plist, cron jobs, or Windows scheduled task running as SYSTEM. Running the watchdog as root/SYSTEM and adding shell aliases are significant privileges and should be granted deliberately. always:false mitigates forced inclusion, but the skill does request system-level persistence.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install openclaw-watchdog-pro
  3. After installation, invoke the skill by name or use /openclaw-watchdog-pro
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v2.2.0
新增错误模式检测:429限流、5xx错误、认证错误、网络错误、消息投递失败,智能分析错误率、突增、集中度
v2.1.0
WebSocket探针+自动doctor修复
v2.0.0
v2.0: 跨平台 Node.js 重构,自动安装脚本,状态查询,配置管理,优雅关闭
Metadata
Slug openclaw-watchdog-pro
Version 2.2.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 3
Frequently Asked Questions

What is OpenClaw Watchdog Pro?

OpenClaw 跨平台配置备份与网关监控。自动备份 openclaw.json,每分钟检查 gateway 状态,宕机时自动恢复。支持 Linux/macOS/Windows。触发:配置备份、gateway 监控、自动恢复、看门狗、watchdog。 It is an AI Agent Skill for Claude Code / OpenClaw, with 173 downloads so far.

How do I install OpenClaw Watchdog Pro?

Run "/install openclaw-watchdog-pro" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is OpenClaw Watchdog Pro free?

Yes, OpenClaw Watchdog Pro is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does OpenClaw Watchdog Pro support?

OpenClaw Watchdog Pro is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created OpenClaw Watchdog Pro?

It is built and maintained by Dorad (@doradx); the current version is v2.2.0.

💬 Comments