Polymer

• Outcome: Review
• Latest audit: May 29, 2026 at 9:35 PM UTC
• Version:
  1.0.1

Overview

This skill is not clearly malicious, but it mixes an unclear Polymer identity with live Membrane actions that can affect connected accounts.

Review before installing. Confirm what Polymer service this connector actually controls, inspect the Membrane permission prompt, and require explicit approval before the agent creates or runs any action that could change records, workflows, or external account data.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Description-Behavior Mismatch

High

• Confidence: 95% confidence
• Finding: The skill metadata says this is a Polymer data integration via Membrane, but the body describes the unrelated Polymer JavaScript UI library and its object model. This mismatch can cause an agent or user to invoke the skill under false assumptions, increasing the chance of unintended authenticated network actions against an external account.

Intent-Code Divergence

High

• Confidence: 96% confidence
• Finding: The contradictory documentation mixes an integration skill with content about the Polymer JavaScript library, creating ambiguity about the skill's purpose and behavior. In an agent setting, ambiguity is dangerous because it can lead to selecting the skill for the wrong task and then executing real operations on a connected third-party account.

Vague Triggers

Medium

• Confidence: 78% confidence
• Finding: The description 'Manage data, records, and automate workflows' is broad enough to match many generic user requests, which can cause over-triggering of a network-capable skill. Over-broad routing increases the risk of unnecessary external actions or disclosure of account-connected data when a narrower, safer tool would suffice.

Missing User Warnings

Medium

• Confidence: 84% confidence
• Finding: The skill tells the agent how to authenticate and connect to an external service but does not prominently warn that commands perform real networked operations against user-linked accounts. Without explicit notice and consent language, users may not understand that listing, creating, or running actions can affect live external systems.

VirusTotal

66/66 vendors flagged this skill as clean.