Tradebot Ops

Monitor trading bot health by detecting stalls, verifying LIVE status and data freshness, restarting safely if needed, and summarizing status clearly.

This skill looks like it was written to operate a real trading bot, but it leaves out key operational details. Before installing, confirm: (1) where the audit logs go and who can read them; (2) whether the agent runtime will have permission to read dist/out/* and to kill/start processes (these are privileged actions); (3) whether the /api/trading endpoints require authentication and, if so, what credentials will be used (the skill declares none); (4) add explicit safety checks and a human-confirm step before automated restarts in production to avoid accidental trading disruption. If you intend to use it, test in a safe staging environment with minimal privileges and explicit, auditable controls for restart actions.

Type: OpenClaw Skill Name: tradebot-ops Version: 1.0.0 The `SKILL.md` file instructs the AI agent to perform operational tasks, including stopping and restarting a trading bot process. Specifically, the instruction "stop bot PID (runtime or heartbeat pid) → apply+restart" implies the agent will execute system commands to manage processes. If the PID is sourced from input files like `dist/out/live_heartbeat_*.json` without robust sanitization, this creates a potential shell injection vulnerability, allowing an attacker who can modify these files to execute arbitrary commands. While the stated purpose is benign (bot operation), this capability introduces a significant security risk due to potential command injection, classifying it as suspicious.