Coding

• Outcome: Review
• Latest audit: May 28, 2026, 11:31 AM
• Version:
  0.1.0

Overview

This skill is transparent about building database-backed dashboards, but it gives agents write access to database structures and recurring external API calls without clear approval or scope limits.

Install only if you want the agent to create database structures, store the data you provide, and build dashboards that poll the listed external API. Before use, specify the exact database or collection, require confirmation before any write, avoid sensitive data unless retention is understood, and verify the endpoint is trusted.

SkillSpector

By NVIDIA

Vulnerability Patterns

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

Findings (5)

Vague Triggers

Medium

• Confidence: 93% confidence
• Finding: The skill description is broadly scoped and can trigger on generic coding requests while authorizing consequential actions like database schema creation, data insertion, and external API integration. Because it lacks explicit activation boundaries, an agent may invoke it in contexts where the user did not clearly consent to persistent writes or network activity.

Vague Triggers

Medium

• Confidence: 94% confidence
• Finding: The instruction block authorizes a wide range of coding and operational behaviors without constraints, including backend design, repeated polling, and database writes. In an agent setting, such open-ended instructions increase the chance of overbroad execution and unintended side effects beyond the user's immediate request.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The skill mandates external API access and 60-second polling without disclosing that it will transmit identifiers and repeatedly contact a remote service. This creates risk of unreviewed data egress, excessive network activity, privacy issues, and operational load if invoked automatically or in the wrong environment.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The skill mandates external API access and 60-second polling without disclosing that it will transmit identifiers and repeatedly contact a remote service. This creates risk of unreviewed data egress, excessive network activity, privacy issues, and operational load if invoked automatically or in the wrong environment.

Missing User Warnings

Medium

• Confidence: 96% confidence
• Finding: The instruction section repeats database-write and continuous polling behavior but omits any warning or approval step for these side effects. In context, this makes the skill more dangerous because the instructions are operationally actionable and could cause persistent changes and repeated outbound requests merely from using the skill.

VirusTotal

66/66 vendors flagged this skill as clean.