ClawHub

Discover

Security checks across malware telemetry and agentic risk

Overview

This skill is a transparent discovery helper that stores local watchlists and optional heartbeat checks, with broad activation language that users should configure carefully.

Install only if you want persistent local discovery memory. Review any proposed AGENTS.md or HEARTBEAT.md changes before approving them, enable heartbeat only for specific topics, and avoid storing secrets or sensitive personal details in ~/discover/.

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
93% confidence
Finding
The routing trigger is broad enough to activate on common requests like 'new ideas' or 'new angles,' which can cause the discover skill to run in situations the user may not intend as persistent discovery behavior. Because the block also instructs reading and writing durable files, an overbroad match can lead to unnecessary persistence, scope creep, and surprising access to stored watchlists or memory.

Vague Triggers

Medium
Confidence
86% confidence
Finding
The detection trigger list includes broad natural-language phrases like 'What else should I know here?' that can match many ordinary conversations and may cause the skill to activate when the user did not intend durable monitoring or discovery behavior. In this skill, over-triggering matters because activation can lead to persistent local state creation, proposed workspace routing changes, and eventual recurring heartbeat behavior, making accidental invocation more security- and privacy-relevant than a normal suggestion heuristic.

Vague Triggers

Medium
Confidence
88% confidence
Finding
The setup instructions define activation triggers such as 'new angles, opportunities, or things they may not know yet' and 'hints at an open loop' in very broad terms, which can cause the skill to engage without clear user consent. In a discovery-oriented skill, this increases the chance of unsolicited persistence, over-collection of user context, and unexpected proactive behavior, especially when combined with later state storage and heartbeat concepts.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal