ClawHub
Back to skill

Security audit

Agent Andri

Security checks across malware telemetry and agentic risk

Overview

This skill is a simple status reporter that appends a fixed idle message to a local meeting-room file, and that behavior is disclosed.

Install only if you want a worker-style status reporter that appends to the meeting-room file every 30 seconds. Treat that file as shared with other skills or agents, stop the script when reporting is no longer needed, and do not provide NV_API_KEY for this version because the script does not use it.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The skill explicitly documents writing status data into another skill's workspace file (`meeting-room/to_leader.txt`), which is a cross-skill data modification boundary. Even if intended for coordination, this creates an integrity and data-flow risk because one skill can alter another skill's inputs or state without an explicit user warning, permission model, or validation boundary.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.