Creator Course Outline

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

Overview

This skill is a local course-outline generator with a disclosed Python helper; it contains unused audit/scanning code, but the normal configured workflow does not activate it.

Reasonable to install for drafting creator-course outlines. Run the helper only on course material you intend to process, choose output paths deliberately because it can create or overwrite files, and review future updates carefully if the unused audit/scanning code becomes exposed or enabled.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• 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
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration

Findings (4)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 92% confidence
• Finding: The skill declares no permissions, yet its instructions explicitly reference reading local resources, writing output files, and invoking `python3`, which implies file and shell capabilities. This is dangerous because callers and policy engines may treat the skill as low-risk while it can access the filesystem and execute local code, increasing the chance of unintended data exposure or command execution.

Description-Behavior Mismatch

High

• Confidence: 95% confidence
• Finding: The dispatcher exposes multiple operational modes such as directory, CSV, pattern, and skill auditing that are unrelated to the declared purpose of generating creator course outlines. This broadens the skill into a general local inspection tool, increasing the chance that an orchestrating agent will use it to enumerate files, inspect repositories, or analyze sensitive local content outside the user’s expected scope.

Context-Inappropriate Capability

High

• Confidence: 96% confidence
• Finding: The pattern scanning mode searches local files for secrets, private URLs, and risky shell constructs, which is a security-audit capability not justified by a curriculum-authoring skill. In an agent setting, this can be abused to inspect arbitrary user repositories or documents and surface sensitive information, even if partially redacted in output.

Context-Inappropriate Capability

Medium

• Confidence: 91% confidence
• Finding: The skill audit logic performs repository completeness and frontmatter compliance checks for skill packages, which is materially different from course-outline generation. That mismatch can cause over-privileged use of the tool against local project directories, exposing structure and metadata that the user did not intend to share through a course-authoring workflow.

VirusTotal

65/65 vendors flagged this skill as clean.