Afrexai Business Automation TEMP

• Outcome: Review
• Latest audit: May 29, 2026 at 9:26 PM UTC
• Version:
  1.0.0

Overview

This skill is mostly a documentation-only automation guide, but it warrants review because it can steer agents toward payment, account, publishing, and recurring-job automations while the package identity is inconsistent.

Before installing, verify that the package you intend to install is the same one being reviewed and that you trust the publisher. Use this skill for planning by default; only let it create scripts, cron jobs, API integrations, payment/account workflows, or public communications after explicit approval, sandbox testing, least-privilege credentials, logging/redaction choices, and a rollback or disable plan are in place.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• 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
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (3)

Missing User Warnings

Medium

• Confidence: 84% confidence
• Finding: The README explicitly encourages implementation of automations via cron jobs, APIs, scripts, and notifications, but does not warn that these actions can execute code, modify external systems, process sensitive business data, or create side effects. In a skill designed to turn an agent into an automation architect, omission of safety boundaries increases the chance that a user or agent will perform impactful actions without confirmation, scope limits, or review.

Missing User Warnings

Medium

• Confidence: 91% confidence
• Finding: The skill provides implementation templates that perform API calls, logging, and file writes for business automation, but it does not consistently warn that these patterns may process sensitive business, financial, HR, or customer data. In this context, users may copy these examples directly into production and inadvertently transmit or persist confidential data without review, minimization, redaction, or access controls.

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: The workflow examples include high-impact actions such as invoice approval, payment scheduling, account provisioning, ticket handling, and content publication, but they do not consistently require human approval before production use. Even though later sections mention human-in-the-loop design, the examples themselves normalize autonomous execution of financially, operationally, and customer-impacting actions, increasing the risk of unsafe deployment.

VirusTotal

65/65 vendors flagged this skill as clean.