---

name: shell-security-ultimate version: 2.2.1 description: "Classify every shell command as SAFE, WARN, or CRIT before your agent runs it. Built for the TinkerClaw fork — github.com/globalcaos/tinkerclaw." metadata: openclaw: owner: kn7623hrcwt6rg73a67xw3wyx580asdw category: security tags: - shell - command-classification - risk-management - agent-safety license: MIT notes: security: "Instruction-only skill that classifies shell commands before execution — it does NOT execute commands itself. Teaches the agent to label every command as SAFE, WARN, or CRIT and enforce approval gates. No binaries, no network calls, no credentials. All logic runs within the existing LLM context."

Shell Security Ultimate

Your agent has root access. Every command it runs is one bad inference away from rm -rf / or curl | bash from a stranger's repo.

This skill won't let that happen.

How It Works

Every shell command gets classified before execution:

• 🟢 SAFE — Read-only, harmless. Runs without friction.
• 🟡 WARN — Could modify state. Logged, flagged, your call.
• 🔴 CRIT — Destructive or irreversible. Blocked until you say so.

No command runs unclassified. No silent chmod 777. No quiet dd if=/dev/zero. Your agent won't accidentally email your SSH keys, won't helpfully format a disk, and won't DROP TABLE users because it misread the task.

What You Get

• Pre-execution classification for every command, every time
• Detailed operation logs so you see exactly what ran and why it was allowed
• Full override control — approve, deny, or escalate at any level

Who It's For

Anyone giving an AI agent shell access and wanting to sleep at night.

Clone it. Fork it. Break it. Make it yours.

👉 Explore the full project: github.com/globalcaos/clawdbot-moltbot-openclaw