Clipboard Knowledge Capture

• Outcome: Pass
• Latest audit: May 29, 2026, 8:21 PM
• Version:
  1.0.0

Overview

This is a local clipboard-to-note helper with disclosed Python file handling and no evidence of hidden network access, credential theft, destructive behavior, or automatic persistence beyond user-chosen output files.

Install only if you are comfortable with a local Python helper reading the input you choose and optionally writing a Markdown or JSON file. Redact secrets and sensitive clipboard contents before use, and avoid modifying the bundled spec to enable audit modes unless you intentionally want local files inspected.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands

Findings (5)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 88% confidence
• Finding: The skill declares no permissions, yet its documented behavior includes invoking python3, reading local resources, and writing output files. This creates a capability-transparency gap: users and policy systems may believe the skill is passive while it can access the filesystem and execute shell-mediated actions.

Intent-Code Divergence

Medium

• Confidence: 83% confidence
• Finding: The skill claims a 'default read-only' security boundary while also documenting execution of a script that produces an output file. Such contradictory safety claims can mislead users into approving actions they would otherwise review more carefully, especially when local knowledge-base writes are involved.

Description-Behavior Mismatch

High

• Confidence: 97% confidence
• Finding: The script's primary behavior is a generic report/audit engine selected by spec mode, not a clipboard-to-local-knowledge capture workflow as declared in the skill metadata. This mismatch is dangerous because it can be used to inspect arbitrary files and directories under the guise of a benign knowledge-capture skill, expanding access and user trust beyond the stated purpose.

Context-Inappropriate Capability

High

• Confidence: 95% confidence
• Finding: The code performs filesystem enumeration, content scanning, and pattern-based detection of secrets and risky commands on arbitrary files or directories. In the context of a clipboard knowledge tool, these capabilities are over-privileged and can expose sensitive local content, making the skill materially more dangerous than advertised.

Vague Triggers

Medium

• Confidence: 87% confidence
• Finding: The trigger examples are broad natural-language phrases that plausibly overlap with ordinary user requests, which can cause the skill to activate when the user did not explicitly intend to invoke it. In this skill’s context, accidental activation could lead to unintended processing and persistence of clipboard-derived content into a local knowledge base, increasing the chance of mishandling sensitive or private material.

VirusTotal

63/63 vendors flagged this skill as clean.