Security audit

Work Productivity Skill Vetter Workflow Helper

Security checks across malware telemetry and agentic risk

Overview

This is a documentation-only workflow helper with overly broad activation wording, but no evidence of hidden code, credential access, persistence, destructive behavior, or data exfiltration.

Install only if you want a general helper for skill-vetting workflows. Be aware it may be invoked too broadly on ordinary security, GitHub, or bug-fix requests, so explicit invocation is safer until the trigger wording is narrowed.

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 (10)

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The trigger phrases are broad enough to match ordinary user requests about security, GitHub, bug fixes, or general productivity, which can cause the skill to activate when the user did not explicitly intend to invoke it. In an agentic environment, over-broad activation can route sensitive or unrelated tasks through this workflow, increasing the chance of unintended actions, misleading outputs, or prompt-scope confusion.

Vague Triggers

Medium

Confidence: 95% confidence
Finding: The trigger phrases are broad enough to match ordinary requests such as asking for help with security, GitHub, bug fixes, or general workflow support. In an agent environment, this can cause unintended activation and route unrelated user tasks into this skill, increasing the chance of inappropriate actions, misleading outputs, or bypass of a more suitable and safer skill selection path.

Vague Triggers

High

Confidence: 96% confidence
Finding: The skill description includes very broad trigger terms such as 'security' and 'first', which are common in ordinary conversation and can cause the skill to activate outside its intended scope. In an agent environment, this can lead to inappropriate routing, unintended instruction injection into unrelated tasks, and reduced trust in skill selection behavior.

Vague Triggers

High

Confidence: 98% confidence
Finding: The explicit keyword list contains ambiguous single-word triggers like 'security', 'first', 'before', and 'github', which are highly likely to match benign user requests unrelated to this skill. Because skills may be auto-selected based on these triggers, this creates a prompt-scope poisoning risk where the skill can hijack unrelated conversations and steer agent behavior unexpectedly.

Vague Triggers

High

Confidence: 95% confidence
Finding: The trigger keywords include very broad, common terms such as “security”, “first”, “before”, and “github”, which can cause the skill to activate in many unrelated conversations. Over-broad activation increases the chance that this skill injects workflow guidance into contexts where it was not requested, creating prompt-routing confusion and potentially interfering with safer or more appropriate skills.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The description says the skill should be used when users ask for broad categories like work-productivity, security, or practical support, but it does not define where the scope ends. This ambiguity can cause accidental invocation for unrelated tasks, reducing predictability and making it easier for the skill to influence conversations outside its intended domain.

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The default_prompt includes very broad trigger terms such as 'help me' and generic work-productivity/security wording, which can cause the skill to activate in unrelated conversations. Because the manifest does not constrain when this prompt should apply, the agent may route ordinary user requests into this skill unexpectedly, increasing prompt-scope confusion and the chance of unsafe or unintended behavior.

Vague Triggers

Medium

Confidence: 93% confidence
Finding: Enabling allow_implicit_invocation without trigger limitations allows the system to invoke this skill automatically based on loosely related language. In combination with the broad default prompt and generic 'security'/'workflow' scope, this creates an overreach risk where unrelated user tasks may be intercepted, leading to unintended actions, confusing outputs, or bypass of user expectations about which skill is operating.

Vague Triggers

High

Confidence: 94% confidence
Finding: The trigger examples are overly broad and include common words and fragments such as 'security' and 'first', which can cause the skill to activate in many unrelated contexts. In a vetting/security-oriented skill, accidental invocation can misroute user requests, override more appropriate skills, or create unsafe trust in irrelevant security guidance.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The activation scope is not clearly bounded: the keyword list and trigger sentences describe when the skill might run, but not when it should abstain. This ambiguity increases the chance of false-positive activation, especially because the skill is framed around practical workflows and security vetting, where misapplication can produce misleading or overconfident outputs in unrelated tasks.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.