Security audit

Word Docx Formatting Repair Helper

Security checks for vulnerabilities and agentic risk

Overview

This is a document-formatting helper with broad activation wording, but it does not include executable code, persistence, credential access, or hidden behavior.

This skill is reasonable to install for Word/DOCX formatting help, but its broad implicit triggers may make it appear for general productivity requests. Before applying any generated find-and-replace or formatting automation to important documents, use a copy, preview the scope, and keep version history or backups available.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

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 (11)

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The trigger sentences are broad, templated, and partially truncated, making activation criteria ambiguous and easier to invoke unintentionally. In an agent ecosystem, overly generic triggers can cause the skill to be selected for requests outside its intended scope, which can lead to incorrect workflow guidance, prompt-routing confusion, or increased exposure to adversarial prompt injection from unrelated tasks.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 87% confidence
Finding: The README promotes document-formatting automation and find-and-replace workflows without warning that these actions can alter, corrupt, or irreversibly overwrite document content and formatting. In this context, users may apply suggested steps directly to important work documents, increasing the risk of accidental data loss or large-scale formatting damage.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The trigger phrases are very broad and map to common productivity requests, which can cause the skill to activate outside the author's intended scope. In an agent environment, overbroad activation increases the chance of unsolicited workflow guidance or document-changing suggestions being injected into ordinary user conversations.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The trigger examples use very generic, everyday help-seeking phrasing that overlaps with normal user requests about workflows or Word documents. This can cause the skill to activate when not specifically intended, increasing the chance of unnecessary context injection, misrouting, or priority hijacking over more appropriate skills.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: 该文件整体为中文本地化版本（文件名为 zh-CN，内容也全部以中文呈现），但文中没有说明这是用户可选的语言版本，也没有提供语言选择或切换指引。按规则，若技能对语言/区域有强制约束而缺少用户选择或明确、合理的区域性说明，应视为自然语言政策问题。
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The skill description and usage conditions are broad enough to match common workplace phrasing, which can cause the skill to activate outside the author's intended scope. Over-broad routing is dangerous because it can mis-handle unrelated user requests, bypass more appropriate skills, or expose users to incorrect automation guidance in unintended contexts.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The trigger keyword list contains generic, high-frequency terms like 'automation' and broad office-productivity language, making accidental invocation likely. This increases the chance of incorrect skill selection, prompt-routing confusion, and downstream unsafe or irrelevant actions based on the wrong workflow.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 89% confidence
Finding: The default prompt uses broad, natural phrasing around common productivity tasks, which overlaps heavily with everyday requests. In combination with implicit invocation, this makes accidental or overly broad skill selection more likely, potentially causing the system to inject this skill into unrelated conversations and increasing the attack surface for prompt-routing abuse.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The skill enables implicit invocation but defines no narrow trigger conditions or guardrails in this file. That means ordinary user requests about Word, formatting, or document editing could automatically route into the skill unexpectedly, increasing the chance of overbroad activation, prompt-surface expansion, and misuse in contexts the user did not clearly intend.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 81% confidence
Finding: The evidence and trigger context prominently include Chinese-language content and titles, but the file does not state that the skill is region-specific or that users can choose their preferred language or locale. For a general productivity skill, this can imply a locale bias without documented justification or opt-in.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The trigger examples are broad enough to match ordinary requests about help, workflows, or Word formatting, which can cause the skill to activate outside a clearly scoped context. Over-broad activation increases the chance of inappropriate routing, unintended prompt injection surface exposure, and user confusion about why this specific skill was invoked.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.