Skill Builder

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only skill-design helper, with the main caution being that its trigger-writing advice can be a bit broad if copied carelessly.

Install this only if you want help designing or auditing OpenClaw skills. When using it, review any generated trigger descriptions so they remain specific and honest, and test near-miss prompts to avoid skills activating on unrelated work.

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: 88% confidence
Finding: This guidance encourages authors to broaden trigger descriptions beyond precise scope, which increases the chance that the skill will activate on adjacent or unrelated requests. In a skill-selection system, overly broad activation criteria can cause misrouting, unnecessary delegation, and instruction collisions that degrade safety and reliability.

Vague Triggers

Medium

Confidence: 92% confidence
Finding: Telling authors to make descriptions 'slightly aggressive about when to fire' directly promotes ambiguous activation behavior. In the context of a skill-builder skill, this is more dangerous because it can systematically propagate overbroad trigger logic into many downstream skills, amplifying selection errors across the ecosystem.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal