Poem Generator

Security checks across malware telemetry and agentic risk

Overview

The advertised poetry skill includes an unrelated marketing script that is not clearly disclosed and can store command text locally.

Review before installing. The poetry helper itself appears low impact, but the package should remove or clearly disclose the unrelated marketing script and its local history logging. Avoid entering private drafts, personal text, or proprietary content unless you have verified which script will run and are comfortable with local persistence.

SkillSpector

By NVIDIA

Vulnerability Patterns

MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
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 (4)

Description-Behavior Mismatch

High

Confidence: 98% confidence
Finding: The script’s documented behavior and command set implement a generic content-marketing assistant, while the skill metadata claims it is a poetry generator. This kind of capability mismatch is dangerous because it misleads users and security reviewers about what the skill actually does, increasing the chance that inappropriate or undisclosed behavior is trusted and executed under false pretenses.

Description-Behavior Mismatch

High

Confidence: 99% confidence
Finding: The actual command outputs are for headlines, SEO, hooks, CTAs, and content metrics rather than poem creation, translation, or literary formatting. In a security context, this is a true integrity and trust violation: users invoking a poetry tool may unknowingly run an unrelated assistant, which can mask broader unauthorized functionality or bypass scrutiny tied to the declared use case.

Intent-Code Divergence

Medium

Confidence: 94% confidence
Finding: The inline documentation explicitly identifies the tool as a content creation and optimization assistant, directly contradicting the published poetry-generator description. This inconsistency is a meaningful security signal because deceptive or inaccurate documentation reduces transparency, hinders auditing, and can conceal unexpected behavior behind a benign-seeming label.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The logging helper writes user-supplied content to a local history file without clear disclosure, consent, retention policy, or an opt-out. In this skill context, prompts may contain sensitive creative drafts or personal text, so silent persistence creates privacy risk and may expose data to other local users, backups, or later compromise of the host.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal