Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
OpenClaw Guardian CN
v1.0.1OpenClaw 系统守护 skill。每日定时自检、自动检测异常并尝试自救。 用于：(1) 每日定时检查 Gateway 状态 (2) 自动检测进程/WebSocket/插件异常 (3) 自动尝试恢复服务 (4) 用户可自定义自检时间 当用户要求设置每日自检、自助恢复、系统守护、自动重启 Gateway 时触发。
⭐ 0· 82·0 current·1 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
Name/description describe a guardian that checks Gateway, processes, WebSocket/plugin status and tries to recover; the scripts and SKILL.md implement those checks (pgrep, openclaw gateway status/start/restart, checking ~/.openclaw, crontab/tasks). Required items (none) are consistent with this purpose.
Instruction Scope
Instructions and scripts read user-local configuration (e.g., ~/.openclaw/openclaw.json, ~/.openclaw/skills/, extensions), run commands (pgrep, openclaw gateway start/restart, node ./dist/daemon-cli.js if present) and propose adding cron entries or Windows scheduled tasks. This is within scope for a guardian, but the SKILL.md includes one-liners that will modify the user's crontab / scheduled tasks if executed — review those commands before allowing automatic execution.
Install Mechanism
No install spec; the skill is instruction- and script-based. No external downloads or obscure URLs are used. Scripts are provided in the package and will not fetch remote archives by the skill itself.
Credentials
The skill does not request environment variables or credentials. It relies on standard environment values (HOME, OSTYPE) and existing local tools (pgrep, node, python3 if present). No unrelated secrets are requested.
Persistence & Privilege
always:false and model invocation defaults are normal. However, the SKILL.md explicitly instructs how to (and provides snippets that will) add cron entries or register Windows scheduled tasks to run guardian checks. Those actions persist on the system and will cause repeated execution if you run the auto-configuration snippets — you should review/approve any crontab/Task Scheduler changes before applying them.
Assessment
This skill appears coherent for an OpenClaw Gateway watchdog: it inspects ~/.openclaw, checks processes, and can attempt to restart the gateway. Before installing or running automated setup: 1) Inspect the provided scripts (check.sh, guardian-daily.sh, heartbeat-check.sh) to confirm you accept their behavior. 2) Do not blindly execute the SKILL.md's auto-configuration one-liners — they will add cron entries or register scheduled tasks that cause persistent automated actions. 3) Understand that the cron entry runs a node start command from your ~/.openclaw directory; if those files are untrusted or modified, that cron could execute arbitrary code. 4) If you want persistent supervision, prefer adding the cron/task manually after review or run the checks interactively first. 5) Note small bugs in scripts (e.g., locale assumptions when parsing free output) — test in a safe environment. If you need, I can point out exact lines to modify or create a safer, minimal version that only reports problems without auto-modifying crontab/tasks.Like a lobster shell, security has layers — review code before you run it.
guardianvk970xrtn28hv05jbd9239knqxh83jm2nhealthvk970xrtn28hv05jbd9239knqxh83jm2nlatestvk9752re7xjjdz73h6q11ce5kg583kqxfsystemvk970xrtn28hv05jbd9239knqxh83jm2n
License
MIT-0
Free to use, modify, and redistribute. No attribution required.