Security audit

Agentvibes Content Skill

Security checks across malware telemetry and agentic risk

Overview

This appears to be an unfinished API-wrapper skill template with unclear activation rules and unnecessary shell-command authority.

Review before installing. This skill looks like an unfinished template rather than a bounded integration: require the publisher to define the actual API, trigger phrases, data sent externally, and remove or tightly justify exec-based diagnostics before using it with sensitive data or broad agent permissions.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands

Findings (6)

Context-Inappropriate Capability

Medium

Confidence: 93% confidence
Finding: The manifest enables the exec tool even though the skill is described as a simple API wrapper, creating unnecessary command-execution capability. This expands the attack surface: ambiguous or malformed prompts could cause the agent to run shell commands unrelated to the stated purpose, especially since later sections explicitly recommend command-line diagnostics.

Intent-Code Divergence

Low

Confidence: 90% confidence
Finding: The error-handling guidance tells the agent to run ping commands even though the skill is presented as an API encapsulation tool returning structured results. That mismatch can normalize shell execution during routine failures and may lead to command execution in contexts where only API interaction was expected.

Intent-Code Divergence

Low

Confidence: 91% confidence
Finding: This later error-handling section again directs command execution for connectivity checks, reinforcing behavior outside the skill’s advertised scope. Repeated instructions to execute network commands increase the chance that an agent with exec permission will perform unnecessary shell actions during error recovery.

Vague Triggers

High

Confidence: 95% confidence
Finding: The trigger text is left as a placeholder, so the skill’s activation boundary is undefined. In an agent environment, ambiguous invocation conditions can cause the skill to activate for unintended requests and route user data to external APIs or privileged tools without clear user intent.

Vague Triggers

High

Confidence: 95% confidence
Finding: The applicable-scenarios section repeats placeholder trigger phrases instead of specifying real usage boundaries. Because the skill also advertises external API usage and includes exec capability, unclear scope materially raises the risk of accidental invocation in unrelated conversations.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: The skill states that it is an API wrapper but does not warn that user-provided data may be transmitted to external services. Without a disclosure boundary, users and orchestrators may unknowingly send sensitive content off-platform, creating privacy, compliance, and data-handling risks.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.