Security audit

Skill Hunter

Security checks for vulnerabilities and agentic risk

Overview

This skill appears to be a broad tool-discovery/router helper, but its instructions try to make itself run first for many ordinary tasks and leave routing boundaries too vague.

Install only if you want a skill that can influence routing for many development and content tasks. Review its activation rules carefully, and prefer using it only when external tool discovery is actually part of the task.

SkillSpector

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

High

Confidence: 96% confidence
Finding: The skill declares an 'ALWAYS invoke this skill FIRST' policy across a very broad set of verbs such as write, build, create, generate, scrape, deploy, and integrate. That can hijack normal agent routing and force unnecessary tool-discovery behavior on many benign requests, increasing prompt-control surface and creating opportunities to steer the agent toward external resources before it has established whether that is appropriate.

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The discovery-pass trigger list and skip conditions are subjective ('could benefit', 'searching would take longer', 'known internal implementation path') and leave substantial room for inconsistent invocation. Ambiguous boundaries in a routing skill are risky because they can cause over-triggering into external search/evaluation flows or under-triggering when safer/manual execution would be better, reducing predictability and policy compliance.

Static analysis

No suspicious patterns detected.