Clear Writing

Security checks across malware telemetry and agentic risk

Overview

This is a documentation-only writing aid with ordinary install instructions and no evidence of hidden execution, data theft, persistence, or privilege abuse.

Safe to install for ordinary writing help. Prefer the registry install or verify the GitHub source before using the GitHub npx command, and avoid sending confidential drafts or secrets through the optional subagent copyediting workflow unless that is acceptable in your environment.

SkillSpector

By NVIDIA

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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: 91% confidence
Finding: The activation guidance is extremely broad: it applies to nearly any task involving human-readable text, including documentation, UI copy, reports, explanations, and editing. In an agentic system, this can cause the skill to be invoked too often, overriding more specialized skills or injecting unnecessary instructions into unrelated workflows, which increases the chance of prompt-scope confusion and unintended behavior.

Natural-Language Policy Violations

Medium

Confidence: 95% confidence
Finding: The text explicitly instructs writers to use the masculine pronoun 'he' for generic antecedents such as 'everybody' or 'anyone.' In a writing skill, this can propagate exclusionary or biased language into generated outputs, creating reputational, accessibility, and policy-compliance risks even though it is presented as style guidance rather than an overtly malicious instruction.

VirusTotal

61/61 vendors flagged this skill as clean.

View on VirusTotal