CashFlow Lite: Self-Sustaining AI Agent

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:35 PM UTC
• Version:
  1.0.0

Overview

This is a text-only money-making planning skill with vague autonomy language, but it does not install code, request credentials, or provide hidden execution capability.

Use this skill as an advisory planning aid. Before acting on any suggested opportunity, review it yourself and do not allow the agent to spend money, use accounts, contact people, scrape sites, create listings, or make commitments unless you explicitly approve that specific action.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• 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

Findings (3)

Vague Triggers

Medium

• Confidence: 95% confidence
• Finding: The README tells users to "give your agent a goal and let it work" without defining any trigger conditions, scope limits, or approval boundaries. In a skill explicitly framed around finding and executing money-making opportunities, this ambiguity can enable unsafe autonomous behavior, including financially risky, policy-violating, or deceptive actions chosen by the agent without clear user consent or constraints.

Missing User Warnings

Medium

• Confidence: 98% confidence
• Finding: The skill is marketed as one that finds and executes money-making opportunities, but it provides no warning that autonomous execution in this domain may create legal, financial, ethical, or platform-abuse risks. Because the context is profit-seeking with an execution focus, omission of a safety warning makes misuse more likely and more dangerous than in a purely informational or educational skill.

Vague Triggers

Medium

• Confidence: 96% confidence
• Finding: The workflow trigger 'For every task' is dangerously overbroad because it causes the monetization-oriented process to activate on all user requests, regardless of relevance, consent, or safety context. In this skill, that broad trigger is more concerning because the stated goal is to find and execute money-making opportunities, which can push the agent to reframe unrelated tasks toward financial extraction or autonomous opportunistic behavior.

VirusTotal

66/66 vendors flagged this skill as clean.