Security audit

长篇网文写作-免费版

Security checks for vulnerabilities and agentic risk

Overview

This fiction-writing skill is mostly coherent, but its broad triggers and automatic local write/exec behavior should be reviewed before installation.

Install only if you intend to use it for personal fiction-writing projects. Before using it, keep work inside a dedicated project folder and review any proposed file writes or shell commands, especially for requests that are not clearly about novel outlines, chapters, character cards, settings, or consistency checks.

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Behavioral ASTexec() Call, eval() Call, Dynamic Import
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration

Findings (4)

Intent-Code Divergence

Medium

Confidence: 93% confidence
Finding: The skill explicitly says team collaboration is outside its supported scope, yet the trigger conditions instruct users to invoke it for team collaboration needs. This inconsistency can cause the agent to apply the skill in unintended contexts, increasing the chance of inappropriate file operations, misleading outputs, or overreach beyond the declared safety boundary.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The trigger conditions are broad enough to match generic project management, task planning, progress tracking, and team collaboration requests, even though this is a fiction-writing tool. Overbroad routing criteria can cause accidental invocation in unrelated workflows, which is risky because the skill has read/write/exec tool access and may create or modify local files outside the user’s intended task.

Vague Triggers

Medium

Confidence: 91% confidence
Finding: The skill states that natural-language instructions can automatically trigger operations, but it does not define clear scope boundaries, approval requirements, or guardrails for what actions may run. In a skill with local read/write/exec capabilities, ambiguous natural-language activation raises the risk of unintended actions and unsafe interpretation of user prompts.

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: The description highlights productivity features but does not clearly warn that the skill manages local files and may create or modify project directories and documents. Users may invoke it without understanding that it can perform persistent filesystem changes, which is particularly concerning given the declared read/write/exec tool set.

Static analysis

No suspicious patterns detected.