Idea To Post

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:22 PM UTC
• Version:
  0.1.0

Overview

This is a markdown-only writing skill that asks guided questions and uses disclosed web research to turn ideas into social posts, with privacy caveats but no hidden or destructive behavior.

Install if you want a guided, research-assisted content workflow. Avoid using it for confidential drafts, client material, unpublished business plans, or sensitive personal stories unless you are comfortable with web search or explicitly disable/decline search in your agent.

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 (4)

Vague Triggers

Medium

• Confidence: 93% confidence
• Finding: The README advertises a very broad trigger phrase—"I want to expand an idea..."—for a general-purpose writing skill. Because this overlaps with normal conversation, the skill may activate unintentionally on benign user messages and steer the session into a multi-turn workflow the user did not explicitly request. In an agent environment, overly broad invocation increases the risk of unintended tool use, data handling, or context capture beyond user expectation.

Missing User Warnings

Medium

• Confidence: 96% confidence
• Finding: The README states that the skill will "automatically search" for supplemental materials and perform "initial" and "precision" external searches, but does not prominently warn users that their topic or derived context may be sent to external tools or services. For a writing-assistance skill, users may provide drafts, business ideas, or sensitive context, so undisclosed search behavior creates a meaningful privacy and data-exposure risk. The content-generation context makes this more dangerous because users are especially likely to share unpublished or proprietary ideas.

Vague Triggers

Medium

• Confidence: 91% confidence
• Finding: The trigger phrases are broad enough to match many ordinary writing or brainstorming requests, which can cause the skill to activate outside its intended scope. That increases the chance users are funneled into a long, tool-using workflow with probing questions and web lookups they did not explicitly request, creating unnecessary privacy and consent risk.

Missing User Warnings

Medium

• Confidence: 96% confidence
• Finding: The skill instructs the agent to automatically perform web searches and integrate external materials before and during questioning, but it does not require a user-facing notice that the user's topic or details may be sent to external services. In a writing-assistance context, users may provide sensitive drafts, business plans, or personal experiences, so silent external querying can leak confidential information and create avoidable data-sharing exposure.

VirusTotal

65/65 vendors flagged this skill as clean.