Back to skill

Security audit

Content Creation Helper

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Chinese social-platform finance content helper with no evidence of hidden access, persistence, data collection, or unsafe installation behavior.

Installers should expect this skill to steer output toward Chinese finance/trading social content and should invoke it only when that platform and language fit the request; generated market commentary should still be reviewed for accuracy and investment-disclaimer compliance before publishing.

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
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill is written to generate Chinese-platform content and strongly implies Chinese-language output by default, but it does not require explicit user preference for Chinese. This can override the user's language expectations, produce inaccessible output, or cause the agent to transform content into a language/platform style the user did not request.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The trigger condition is broad enough that it could activate for generic requests about writing social posts, even when the user did not ask for finance/trading content or for this specific platform adaptation. That can cause unintended skill selection, leading the agent to steer responses into specialized financial-content generation and platform-specific formatting outside the user's actual intent.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The module description and exposed functions are all tailored to Chinese platforms and Chinese-language output, and downstream logic hardcodes Chinese copy patterns for titles, disclaimers, and formatting. This creates a natural-language locale restriction without any opt-in, configurability, or explanation that the skill is intentionally region-specific.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The optimize_title function automatically rewrites titles into Chinese question forms for Zhihu and adds a Chinese bracketed prefix for WeChat. This is a language-policy issue because the skill imposes a specific language/locale style rather than letting the user choose or clearly declaring a justified regional scope.

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
86% confidence
Finding

The SEO/title optimization trigger is ambiguous because it does not clearly limit itself to finance/trading content or the three named platforms. In practice, this can misroute unrelated SEO requests into this skill, causing scope confusion, inappropriate formatting, or domain-specific framing the user did not request.

Content

No source excerpt is available for this finding.