Word Docx Formatting Repair Helper

• Outcome: Pass
• Latest audit: Jul 29, 2026 · 5:43 AM UTC
• Version: 0.20260729.110214

Overview

This is a documentation-only Word/DOCX formatting helper with overly broad activation wording but no hidden code, credential use, persistence, or destructive behavior.

This skill is reasonable to install for Word or DOCX formatting repair help, but users should be aware it may auto-activate on broad productivity or editing prompts. Prefer invoking it explicitly when working on document formatting, styles, or find-and-replace tasks.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (11)

Vague Triggers

Medium

• Confidence: 86% confidence
• Finding: The trigger sentences are broad, repetitive, and include awkward truncated phrasing that could cause the skill to activate for loosely related requests rather than clear Word/document-formatting tasks. Overbroad activation increases the chance of unintended routing, prompt collisions, or misuse of the skill in contexts it was not designed to handle, which is a genuine security and reliability concern for agentic systems.

Vague Triggers

Medium

• Confidence: 94% confidence
• Finding: The trigger phrases are extremely generic (for example, "Help me" and "I need a practical workflow for ..."), which can cause the skill to activate on ordinary user requests that were not intended for this specific skill. In an agent environment, overbroad invocation increases the chance of misrouting user tasks, unexpected skill execution, and prompt-scope confusion, even though the content here is not overtly malicious.

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: The skill description is broad enough to match many ordinary productivity requests, which increases the chance the agent will invoke this skill outside its intended boundaries. Over-broad routing can cause misapplication of automation guidance, reduce least-privilege behavior in skill selection, and interfere with safer or more specialized skills.

Vague Triggers

Medium

• Confidence: 95% confidence
• Finding: The listed keywords are generic terms like productivity, Word, styles, and automation that commonly appear in unrelated user requests. This makes accidental invocation more likely and can lead to incorrect tool/skill routing, especially in systems that rely heavily on keyword matching.

Vague Triggers

Medium

• Confidence: 94% confidence
• Finding: The example triggers use vague, natural phrasing that overlaps with normal conversation, so they may collide with routine requests not meant for this skill. In an automated dispatch system, such examples can effectively broaden the skill's activation surface and degrade precision of intent classification.

Vague Triggers

Medium

• Confidence: 87% confidence
• Finding: The trigger keywords are broad, generic productivity terms such as 'automation' and 'styles' that commonly appear in ordinary conversation. This can cause the skill to activate outside the author's intended scope, leading to unintended routing of user requests and potentially exposing the model to irrelevant or unsafe instruction contexts.

Vague Triggers

High

• Confidence: 93% confidence
• Finding: The example trigger phrases are extremely permissive and resemble normal help-seeking language, which increases the chance of accidental invocation during unrelated conversations. Because examples often shape downstream trigger matching behavior, this broad wording can materially increase false activations and misapplication of the skill.

Vague Triggers

Medium

• Confidence: 93% confidence
• Finding: The default prompt is phrased so broadly that it can match ordinary user requests about productivity, Word documents, formatting, or editing tasks without a clearly bounded activation condition. This increases the chance of unintended skill invocation, which can cause the agent to inject skill-specific instructions or behavior in contexts where the user did not explicitly request it.

Vague Triggers

Medium

• Confidence: 96% confidence
• Finding: Enabling implicit invocation without specific activation constraints allows the skill to be auto-selected during many normal conversations related to writing or document editing. In combination with the broad prompt, this expands the attack surface for prompt steering, unintended tool usage, or policy bypass through accidental invocation of the skill.

Vague Triggers

High

• Confidence: 91% confidence
• Finding: The trigger sentence is so broad and generic that ordinary user requests about Word, formatting, or general productivity could unintentionally activate this skill. Over-broad activation increases the chance of prompt-routing mistakes, causing the wrong skill to handle unrelated requests and potentially exposing users to irrelevant or lower-quality automation guidance without clear consent.

Vague Triggers

Medium

• Confidence: 87% confidence
• Finding: The invocation examples do not define clear boundaries for when the skill should or should not run, which makes routing ambiguous. In practice this can lead to accidental invocation on loosely related requests, reducing predictability and enabling untrusted skill behavior to be inserted into normal conversations more often than intended.

VirusTotal

64/64 vendors flagged this skill as clean.

Static analysis

No suspicious patterns detected.