ClawHub
Back to skill

Security audit

Discover

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed persistent discovery helper that stores local watchlists and only enables recurring checks after user approval.

Install this only if you want the agent to keep a local discovery watchlist and possibly run approved recurring checks. Review the activation and heartbeat preferences during setup, and avoid adding sensitive topics unless you are comfortable with topic keywords being used in public web searches.

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 (3)

Vague Triggers

Medium
Confidence
89% confidence
Finding
The activation rule uses broad natural-language triggers like 'new ideas', 'new angles', and 'keep discovering this over time', which can match ordinary user requests and cause the skill to activate more often than intended. Because activation also directs the agent to read and write persistent files, accidental triggering can expand data retention and autonomous behavior beyond the user's explicit consent.

Vague Triggers

Medium
Confidence
82% confidence
Finding
The template says discovery should note when it should 'activate automatically' but does not define clear triggers, boundaries, or consent checks. In a memory-driven skill, this ambiguity can cause unsolicited behavior, over-collection of contextual signals, or repeated proactive prompting that the user did not intend.

Vague Triggers

Medium
Confidence
89% confidence
Finding
The activation criteria are intentionally broad ('new angles, opportunities, or things they may not know yet' and open-loop life/work changes), which can cause the skill to engage in many ordinary conversations without clear user intent. In a discovery skill that stores durable watchlists and future behavior, this increases the risk of over-collection, unwanted persistence, and proactive behavior being enabled in contexts where the user only wanted a one-off discussion.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.