Decomposes complex user requests into executable subtasks, identifies required capabilities, searches for existing skills at skills.sh, and creates new skills when no solution exists. This skill should be used when the user submits a complex multi-step request, wants to automate workflows, or needs help breaking down large tasks into manageable pieces.

• Outcome: Review
• Latest audit: May 28, 2026, 4:47 AM
• Version:
  1.0.0

Overview

This skill is a coherent planning helper, but it can lead an agent to search for, globally install, or create other skills without clear install consent or containment.

Install only if you want an agent to help plan workflows and discover skills. Before letting it install or create anything, review each proposed skill source, avoid automatic `-g -y` installs, confirm required credentials or scheduled jobs, and keep a way to remove any added skills.

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 (2)

Missing User Warnings

Medium

• Confidence: 88% confidence
• Finding: The README explicitly promotes networked skill search, downloading/installing skills, and automatic skill creation, but it does not warn users that these actions may transmit prompts/data to external services or modify the local/global environment. In a skill whose purpose is to discover and install code-like extensions, the lack of consent, trust-boundary disclosure, and safety guidance increases the chance of unintended external data exposure or installation of untrusted components.

Vague Triggers

Medium

• Confidence: 88% confidence
• Finding: The skill is described to trigger on very broad conditions such as any complex multi-step request, automation workflow, or help breaking down large tasks. Overbroad activation can cause the agent to invoke a powerful orchestration skill in many ordinary contexts, increasing the chance it will search for external skills, generate executable plans, or propose new automation without sufficiently narrow user intent. In this file, that broad scope is made more sensitive by later instructions to search/install third-party skills and create new ones.

VirusTotal

64/64 vendors flagged this skill as clean.