Plan

Security checks across malware telemetry and agentic risk

Overview

This skill appears to provide planning and self-improvement guidance for the agent without requesting system access, credentials, persistence, or external actions.

Safe to install for improving agent planning behavior. Be aware that its broad description may make it activate often, so users who prefer minimal planning overhead may want narrower activation wording, but there is no evidence of unsafe system access or hidden behavior.

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 (1)

Vague Triggers

Medium

Confidence: 95% confidence
Finding: The manifest description says the skill 'Auto-learns when to plan vs execute directly' and 'Adapts planning depth to task type,' but it does not define explicit trigger phrases, boundaries, or exclusion conditions for when this skill should activate. In a markdown/manifest-scoped file, this broad wording could cause the skill to match many ordinary requests that involve doing work, making invocation behavior unclear.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal