Note Taker

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:18 PM UTC
• Version:
  2.0.1

Overview

This is a locally run note/task helper with confusing packaging, but the reviewed artifacts do not show hidden, networked, privileged, destructive, or credential-seeking behavior.

Install only if you are comfortable with a mixed note-template and task-list package. Avoid storing secrets or highly sensitive notes in the task script because it keeps local data and command history until you delete them, and confirm which script your installed command runs.

SkillSpector

By NVIDIA

Vulnerability Patterns

• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (5)

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 96% confidence
• Finding: The skill metadata and title present this as a note-taking assistant, but the documented behavior is a persistent task manager that stores data locally, tracks history, exports data, and manages reminders and priorities. This mismatch is dangerous because users, reviewers, or policy systems may grant permissions or trust assumptions appropriate for note formatting while the skill actually handles broader stateful task-management functions and data persistence.

Description-Behavior Mismatch

High

• Confidence: 98% confidence
• Finding: The manifest advertises note-taking and note-organization, while the body documents a to-do and planning tool with reminders, priorities, weekly views, and persistent logs. Security-wise, this is an integrity and transparency problem: downstream tooling or users may approve the skill under one capability profile while it operates under another, increasing the chance of unintended data collection, storage, or use.

Intent-Code Divergence

Medium

• Confidence: 92% confidence
• Finding: The file title and manifest claim a note-taking skill, but the embedded documentation immediately reframes it as a productivity and task-management tool. While not an exploit by itself, this contradiction weakens trust boundaries and can conceal the real operational scope of the skill, especially because it persists data and command history.

Description-Behavior Mismatch

Medium

• Confidence: 97% confidence
• Finding: The skill metadata says this is a note-taking assistant for Cornell notes, Zettelkasten, mind maps, meeting notes, and lecture notes, but the script actually implements a generic todo/task manager. This kind of capability mismatch is dangerous because users and orchestrators may grant trust, permissions, or invoke the skill under false assumptions, resulting in deceptive behavior and unintended data handling.

Intent-Code Divergence

Medium

• Confidence: 95% confidence
• Finding: The inline help and comments describe the tool as a productivity and task management utility, directly contradicting the manifest's note-taking description. Conflicting declarations increase supply-chain and agent-integration risk because reviewers or automated systems may rely on one description while the code behaves according to another.

VirusTotal

38/38 vendors flagged this skill as clean.