Security audit

AI写作提示词与公众号爆款文案指南

Security checks for vulnerabilities and agentic risk

Overview

This is a markdown-only Chinese social-media writing guide with no code or persistence, but it explicitly teaches full reproduction and publish-ready rewriting of third-party platform content.

Review this skill carefully before installing. It appears technically low-risk, but users should avoid using its extraction workflow to copy full third-party articles, posts, subtitles, or scripts; use it for original drafting, structure analysis, short excerpts, or attribution-safe transformation instead.

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

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
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

Findings (3)

Vague Triggers

Medium

Confidence: 95% confidence
Finding: The manifest description says the skill should be used when helping users 'create writing prompts,' 'generate article drafts,' 'rewrite content,' or 'create click-worthy titles.' These are very broad, everyday writing-assistance requests with no clear boundaries or exclusion conditions, so the skill could be invoked unintentionally for generic writing tasks.

Natural-Language Policy Violations

Medium

Confidence: 90% confidence
Finding: The skill title and all operational instructions are written exclusively in Chinese, and one embedded prompt explicitly mandates '使用自然、流畅有节奏感的中文.' This imposes a specific language/locale on outputs without opt-in or explanation, which is a natural-language policy concern under the locale-choice rule.

Context-Inappropriate Capability

Medium

Confidence: 95% confidence
Finding: This section explicitly instructs the model to perform '100% pure restoration' of complete third-party platform copy, including titles, body text, emojis, hashtags, and spoken scripts. In the context of a writing-assistance skill, that enables large-scale reproduction of copyrighted content and plagiarism rather than transformative assistance, making misuse straightforward.

Static analysis

No suspicious patterns detected.