Aibrary 100

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:15 PM UTC
• Version:
  0.1.0

Overview

This is a simple instruction-only book recommendation skill with no code, credentials, or system access requested.

Safe to install as a curated reading-list helper. Be aware it may answer some generic book-recommendation prompts with the Aibrary framework unless you ask for a different source or style.

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 (1)

Vague Triggers

Medium

• Confidence: 87% confidence
• Finding: The trigger description is broad enough to activate on generic book-recommendation queries such as 'must-read books' or 'essential reading list,' which may cause the agent to invoke this skill outside its intended Aibrary-specific context. This is not directly exploitable as code execution or data exfiltration, but it can lead to misrouting, irrelevant responses, and degraded agent reliability.

VirusTotal

66/66 vendors flagged this skill as clean.