ClawHub
Back to skill

Security audit

Auto Research Pipeline

Security checks across malware telemetry and agentic risk

Overview

This skill is not malicious, but one command can trigger broad automated experiments and multi-agent project changes without clear boundaries or approval checkpoints.

Install only if you want a single command to run an automated research and project-delivery pipeline. Use it in a version-controlled or backed-up workspace, constrain the files and commands AutoResearch may touch before running, and review results before allowing downstream agents to apply 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 (1)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill explicitly states that a single invocation will automatically execute four stages, including repeated experimentation and then invoking another orchestration skill in the background, but it does not present a clear user-facing warning about this chained behavior. This is dangerous because users may trigger broad, multi-step actions with side effects, resource consumption, and delegated agent activity without informed consent or an opportunity to review each stage.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.