ClawHub
Back to skill

Security audit

Recipe

Security checks across malware telemetry and agentic risk

Overview

This is a planning-only skill that is mostly coherent and disclosed, with only a minor risk that broad trigger wording could activate it when the user meant a different kind of plan.

Install this if you want an agent to produce very detailed, commit-ready implementation plans from already-approved specs. Be aware it may trigger on generic planning phrases, and it may create and commit a plan file in the repository when used.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Vague Triggers

Medium
Confidence
92% confidence
Finding
The skill description uses broad triggers such as "write the plan" and "plan this out," which can activate the skill for many ordinary user requests outside a tightly scoped implementation-planning context. Overbroad activation can cause the agent to select this skill inappropriately, leading to workflow hijacking, unnecessary instruction precedence, or generation of rigid implementation artifacts when the user intended something else.

VirusTotal

59/59 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.