Using Superpowers Tianjin

• Outcome: Review
• Latest audit: May 29, 2026 at 9:25 PM UTC
• Version:
  0.1.0

Overview

This skill does not run code or access data, but it broadly tries to control the agent’s behavior at the start of every conversation.

Install only if you intentionally want a strict global meta-skill that can make the agent check skills before ordinary answers and even before clarification. It has no code or data-access capability, but it may slow, distort, or over-constrain everyday agent interactions.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• 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 (2)

Vague Triggers

High

• Confidence: 96% confidence
• Finding: The skill declares it should be used when starting any conversation, which creates an excessively broad activation scope for a process-controlling skill. Because it governs behavior before any response, this can force unnecessary tool invocation, override normal task triage, and create a prompt-level denial of service where the agent becomes trapped in recursive or irrelevant skill checks.

Vague Triggers

High

• Confidence: 98% confidence
• Finding: The 'even a 1% chance a skill might apply' rule is vague and effectively mandates invoking skills under near-all circumstances. In context, this is especially dangerous because the skill is itself a meta-instruction about how all future actions must be performed, so the ambiguity can cascade into excessive tool use, recursive skill loading, and suppression of normal reasoning or clarification flows.

VirusTotal

63/63 vendors flagged this skill as clean.