ClawHub

Agent Ai Ml Ops Specialist

Security checks across malware telemetry and agentic risk

Overview

This appears to be a broad MLOps guidance skill with no evidence of hidden execution, credential use, or persistence, though its activation scope is wide.

Install this if you want a broadly invoked MLOps/production ML assistant. Review its trigger wording if your environment has many adjacent data, compliance, or DevOps skills, and require explicit confirmation before applying production, compliance, or infrastructure changes.

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
85% confidence
Finding
The skill description says to use the agent when requests match the domain or role, but it provides no narrowing criteria, exclusions, or thresholds. In multi-skill routing systems, this can cause over-selection of a powerful specialist with broad tooling, increasing the chance of inappropriate invocation, unnecessary tool use, and expanded attack surface from loosely related prompts.

Vague Triggers

Medium
Confidence
92% confidence
Finding
The auto-trigger list covers a very wide set of ML topics, incident types, platforms, and compliance areas without constraints, priority rules, or disambiguation from adjacent skills. This broad matching can let unrelated or marginally related requests activate the skill, and because the imported spec references extensive tools and operational workflows, mistaken invocation could lead to overreach, unsafe actions, or leakage through unnecessary access paths.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal