communication-mqtt

Security checks across malware telemetry and agentic risk

Overview

This is a user-directed MQTT helper for sharing agent intro and status messages on a local broker, with privacy caveats but no hidden execution or exfiltration behavior.

Use this only with a trusted local MQTT broker. Avoid publishing secrets or sensitive work details in agent IDs, roles, or activity fields, treat subscribed messages as untrusted text, and clear retained MQTT topics if old status data should not remain visible.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
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: 93% confidence
Finding: The skill explicitly instructs users to publish and subscribe to agent identity and status data over an MQTT broker, but it does not warn that these messages may expose operational metadata, identifiers, or activity patterns to other local or network participants. Because MQTT is a network messaging system and the examples encourage sharing agent IDs, roles, and activities, users may unintentionally disclose sensitive information or enable profiling if the broker is shared, misconfigured, or later exposed beyond localhost.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal