ClawHub
Back to skill

Security audit

Reclaim Your Brain

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only learning skill that changes the assistant into a Socratic tutor and does not request system access or handle sensitive data.

Install this if you want the assistant to slow down and teach through questions instead of giving immediate answers. Be aware that broad phrases like asking to understand something deeply may activate the mode, so users who need quick direct answers may need to avoid or disable the skill.

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 (2)

Vague Triggers

Medium
Confidence
90% confidence
Finding
The skill metadata includes broad activation examples such as "make me understand this" and "don't just give me the answer," which are common phrases users may say in many unrelated contexts. This can cause unintended activation and silently alter the assistant's behavior, creating reliability and prompt-routing risk even though the skill's content is educational rather than overtly harmful.

Vague Triggers

Medium
Confidence
88% confidence
Finding
The listed trigger phrases are still broad and conversational, especially patterns like "don't just give me the answer" or "make me understand [topic], don't just tell me." In normal assistant usage these phrases could appear organically, leading to accidental activation and unexpected changes in response style or workflow.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.