aida

Security checks across malware telemetry and agentic risk

Overview

This AIDA skill is purpose-aligned but should be reviewed because it can invoke authenticated smart-building control and optimization endpoints without documented confirmation or safety boundaries.

Install only if you trust the AIDA endpoint and can provide a least-privilege token. Treat control and optimization intents as real building operations, require human confirmation before use, and verify API responses because the current code may report success even when the API call fails.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Vague Triggers

Medium

Confidence: 81% confidence
Finding: The example utterance "Optimize for energy savings." is a broad, high-level command that can trigger authenticated optimization actions without clarifying scope, confirmation, or safety constraints. In a smart-building context, ambiguous optimization requests could cause unintended changes to HVAC, lighting, or other building systems, making the prompt design materially risky.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The skill advertises authenticated control and optimization capabilities over building systems but does not warn users that actions may directly affect physical environments and operational conditions. This omission increases the chance of unsafe or unintended use, especially because the documented intents include device control and building-wide optimization backed by bearer-token-authenticated API calls.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal