Screenshot Capture

Security checks across malware telemetry and agentic risk

Overview

This skill transparently saves user-shared screenshots into local notes and reminders, with privacy cautions but no evidence of hidden, destructive, or exfiltrating behavior.

Install this only if you want screenshot contents saved locally and summarized into notes, with a one-week reminder created by default. Avoid using it on screenshots containing passwords, tokens, private messages, financial data, or proprietary material, and tell the agent when you do not want a screenshot saved or a reminder created.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
89% confidence
Finding
The skill's invocation description is broad enough to activate on ordinary screenshot-sharing and loosely contextual comments like 'save this,' which can cause the agent to perform storage, extraction, categorization, and reminder-setting without sufficiently explicit user intent. In this skill, that over-breadth is more dangerous because activation leads directly to persistent file writes and note-taking actions, increasing the chance of unintended data retention or privacy-sensitive processing.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill instructs the agent to copy inbound images into local storage and update multiple notes files, but it provides no safety checks, user confirmation, path validation, or warning that local files will be modified. This is dangerous because screenshots often contain sensitive or proprietary data, and automatic persistence can create unintended data exposure, overwrite risks, or unreviewed changes to the user's knowledge base.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal