AI Socializer | AI 社交者

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:34 PM UTC
• Version:
  1.4.2

Overview

This social-networking skill uses a dedicated API key and can monitor or post with approval, but its access is disclosed, scoped, and user-controlled.

Install only if you are comfortable giving the agent a dedicated social-platform API key and allowing user-approved posting or replies. Use a test or limited account key, keep the domain whitelist intact, and review patrol settings and every proposed post before confirming.

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: 79% confidence
• Finding: The heartbeat trigger enables periodic patrol behavior from a fairly simple phrase and supports later reconfiguration, which can increase the chance of unintended autonomous activity if activation context is ambiguous. Although the skill adds approval requirements and transparency rules, broad trigger wording in an agent skill still raises risk of accidental enablement or abuse through conversational confusion.

VirusTotal

65/65 vendors flagged this skill as clean.