Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Agent Ops Kit
v1.0.0Production-grade health monitoring, alerting, and service management for OpenClaw agents. Monitor URLs, auto-restart services, get Telegram alerts, track upt...
⭐ 0· 37·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name and description match the included assets: a shell health-check script, a Python health-check class, and a sample JSON config. Required resources (none declared) align with a lightweight, instruction-only monitoring tool.
Instruction Scope
SKILL.md instructs creating ~/.agent-ops, copying/running the provided scripts, scheduling via cron, and placing a Telegram bot token in the config. The runtime instructions and scripts only contact user-configured targets and the Telegram API. However, the scripts execute restart commands taken directly from the config (via eval), which means the tool will run arbitrary shell commands the user places in services.json; that is expected for an auto-recovery tool but is a privileged action and should be constrained and reviewed by the user.
Install Mechanism
Instruction-only skill with bundled scripts — no network downloads or package installs. Files are written only to user home paths (~/.agent-ops). No high-risk install operations detected.
Credentials
The skill does not declare any required environment variables or secrets. It expects Telegram credentials in the user config file; storing tokens in ~/.agent-ops/config/services.json is functional but places secrets on disk in plaintext. Sample restart commands include 'sudo' and systemctl/docker commands — these are proportional to service restart needs but imply the script may require elevated privileges when run, depending on configuration.
Persistence & Privilege
always:false and no system-wide modifications are requested. The skill writes state, logs, and metrics under ~/.agent-ops (normal). Note: the platform-default autonomous invocation is allowed; if the agent is permitted to invoke the skill autonomously, it could trigger configured restart commands — review config and permissions before enabling autnonomous runs.
Assessment
This toolkit appears to do what it says: health checks, metrics, Telegram alerts, and optional auto-restarts via commands you put in services.json. Before installing: 1) inspect and lock down ~/.agent-ops/config/services.json (it will contain your Telegram bot token and restart commands) and set file permissions so others cannot read/modify it; 2) review every restart_command you add — the script will eval and run them (including sudo/systemctl/docker commands); avoid putting arbitrary shell commands there; 3) if you must allow service restarts, prefer restricted sudoers entries so the monitoring user can only run the specific restart commands required; 4) run and test in a non-production environment first to ensure restart behavior and cron scheduling act as you expect; 5) if you enable autonomous agent invocation, be aware the agent could trigger restarts and alerts automatically — only enable that if you trust the agent and the config.Like a lobster shell, security has layers — review code before you run it.
latestvk978bwyxwa5h1qq6a96akcjk0x84rvbp
License
MIT-0
Free to use, modify, and redistribute. No attribution required.