Sophiie AI Office Manager

• Outcome: Pass
• Latest audit: May 27, 2026 at 10:23 PM UTC
• Version:
  0.0.3

Overview

This skill is a clear Sophiie API wrapper that can manage CRM data and send calls or SMS, so it is powerful but not deceptive.

Install only if you intend your agent to manage a live Sophiie account. Use a test or least-privileged API key where possible, and require explicit confirmation before sending SMS or calls, deleting records, or changing customer-facing FAQs and policies.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (3)

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The README advertises actions that can send SMS, place calls, and create, update, or delete business records, but it does not clearly warn that these are real external side-effecting operations. In an agent-skill context, that omission is dangerous because users may invoke natural-language commands assuming they are informational, leading to unintended communications, record changes, or privacy-impacting actions against live customer data.

Vague Triggers

Medium

• Confidence: 76% confidence
• Finding: The skill description and invocation guidance are broad enough that ordinary requests like 'call', 'text', 'show leads', or business-management phrasing could trigger this skill unintentionally. Because the skill supports destructive and externally visible actions such as sending SMS, placing calls, and deleting records, accidental activation can cause real-world side effects and data disclosure.

Missing User Warnings

Medium

• Confidence: 89% confidence
• Finding: The SMS capability is documented without a prominent user-facing warning about external message delivery, recipient impact, or billing/communication consequences. In a skill that can message leads directly, missing warnings increase the risk of accidental outbound communications, privacy violations, and unauthorized contact with customers.

VirusTotal

66/66 vendors flagged this skill as clean.