ClawHub
Back to skill

Security audit

Managed Agent Growth Loop

Security checks across malware telemetry and agentic risk

Overview

The skill appears to be a disclosed workflow helper for agent retrospectives and follow-up planning, with no artifact-backed evidence of hidden execution, exfiltration, or destructive behavior.

Install only if you want an agent to help structure managed-agent retrospectives and follow-up workflows. Before using it, name the client or project, specify what history or notes it may use, and confirm before creating persistent tracking files, reminders, or cron-style follow-up recommendations.

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
95% confidence
Finding
The description embeds many broad natural-language trigger phrases such as '현장 회고', '피드백 루프', and '선제적 행동', which can match routine conversations and cause the skill to activate outside its intended scope. Because this skill proposes operational tracking, reminders, cron promotion, and communication drafts, unintended invocation can lead to unnecessary processing of customer context and accidental generation of follow-up actions.

Vague Triggers

Medium
Confidence
93% confidence
Finding
The execution trigger section describes applicability in very general terms like reviewing usage history, summarizing strengths/limitations, or organizing next steps, without clear guardrails on when the skill should not run. In context, this expands activation into common project-management discussions, increasing the chance the agent will over-collect operational details or initiate workflow changes for the wrong client or task set.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.