Prompt Engineering Expert

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:33 PM UTC
• Version:
  1.0.0

Overview

This is a documentation-only prompt-engineering guide with one questionable refusal-troubleshooting example, but no hidden code, persistence, credential handling, or destructive behavior.

Install this if you want prompt-writing and prompt-debugging reference material. Use the refusal-troubleshooting advice only for legitimate, safe requests, and do not use the skill to rephrase harmful requests to bypass model safeguards.

SkillSpector

By NVIDIA

Vulnerability Patterns

• 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
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (2)

Ssd 1

Medium

• Confidence: 94% confidence
• Finding: The troubleshooting guidance recommends fictional framing as a way to obtain analysis of manipulative behavior after a refusal. Even though presented as a writing or psychology context, this normalizes a refusal-bypass pattern: repackaging harmful intent into a narrative frame to elicit operationally useful guidance about manipulation tactics.

Ssd 2

Low

• Confidence: 90% confidence
• Finding: The advice to 'clarify legitimate purpose,' 'reframe the question,' and 'explain why you need this' can teach users how to paraphrase disallowed requests to get past model refusals. In a prompt-engineering skill, this is more sensitive because the audience is specifically learning techniques to shape model behavior, making safety-bypass patterns more actionable.

VirusTotal

65/65 vendors flagged this skill as clean.